Unlocking superionic conduction by modulating electrostatic interactions in a zirconium-based trigonal halide solid electrolyte

Abstract

Halide solid electrolytes have recently emerged as promising candidates for solid-state batteries, combining high ionic conductivity, favorable cathode compatibility, and mechanical softness. Among them, the zirconium-based halide Li₂ZrCl₆ has attracted attention due to its cost advantages over its rare-metal-based counterparts, but its intrinsic ionic conductivity remains insufficient for practical applications. While aliovalent substitution has been explored to improve performance, the mechanisms governing lithium transport and dopant interactions remain unclear. In this study, we systematically investigate how aliovalent cation substitution modulates lithium diffusion within the trigonal Li₂ZrCl₆ framework. By integrating computational and experimental approaches, we show that reducing the electrostatic repulsion between cations and lithium ions not only facilitates local lithium diffusion near substituted sites but also unexpectedly generates additional diffusion pathways beyond conventional doping effects, resulting in a more robust percolated diffusion network. Building on these insights, we propose a simpler, dopant-free strategy based on vacancy-mediated substitution, yielding Li_2.3Zr_0.925Cl₆ that exhibits rapid hopping kinetics and enables efficient lithium incorporation, resulting in significantly enhanced ionic conductivity along with improved rate capability and stability. This work reveals a clear composition-path dependency on structurally tolerant halide solid electrolytes, offering a viable route toward high-performance, cost-efficient solid electrolytes.

---

Broader context

All-solid-state batteries require safe, cost-effective solid electrolytes to replace conventional flammable liquid systems. Halide solid electrolytes have emerged as highly promising candidates due to their intrinsic combination of excellent oxidative stability, favorable cathode compatibility, and suitable mechanical deformability. Among them, Li₂ZrCl₆ is particularly attractive owing to the Earth-abundant zirconium that provides a substantial economic advantage. However, its practical implementation is bottlenecked by intrinsically low ionic conductivity. While traditional aliovalent doping improves transport, it relies on expensive, scarce elements without a clear understanding of the underlying mechanistic kinetics. In this work, we reveal that reducing the electrostatic repulsive force exerted by metal cations is the fundamental key to enhancing lithium diffusion in the trigonal Li₂ZrCl₆ framework. This weakened electrostatic environment unlocks previously inaccessible face-sharing pathways and induces local lithium accumulation to facilitate concerted migration. Building on these insights, we introduce a “zero-cost” vacancy-mediated substitution strategy (Li_2+4xZr_1−xCl₆) that maximally suppresses electrostatic repulsion. Achieving an enhanced ionic conductivity of 0.8 mS cm⁻¹ and superior rate capability, this study establishes a fundamental design principle for developing economically viable, high-performance Zr-based solid electrolytes for next-generation energy storage.

Introduction

All-solid-state batteries have gained considerable attention as a safer alternative to conventional lithium-ion batteries that pose significant safety risks due to their flammable organic liquid electrolytes.^1–4 However, the practical implementation of all-solid-state batteries has been consistently plagued by persistent interfacial challenges between solid electrolytes and conventional electrodes.^5–8 In this context, halide-based solid electrolytes, particularly chloride variants, have emerged as promising candidates as catholytes due to their intrinsic combination of excellent oxidative stability and suitable mechanical deformability,^9,10 overcoming the respective limitations of inherently brittle oxide- and chemically reactive sulfide-based electrolytes.^11–15 Among the various structures of halide superionic conductors,^16–19 the trigonal framework has emerged as a versatile and promising structural platform capable of accommodating a diverse range of metal cations including Yb, Tb, Dy, Ho, Y, and Er.^16,20–24 Thus, it has not only inspired the continuous exploration of new variants but has also provided an expansive compositional landscape to systematically tune and optimize their properties.^25,26 However, the practical deployment of this family remains constrained, as the accessible compositional space has been predominantly limited by scarce and costly rare-metal cations.

In this respect, Li₂ZrCl₆ has garnered considerable attention, as the Earth-abundant zirconium element provides a significant economic advantage, which can be even more favorable than that of state-of-the-art sulfide electrolytes, considering the elemental cost.²⁷ Furthermore, Li₂ZrCl₆ features a unique combination of desirable properties for a catholyte, including high oxidative stability, mechanical compatibility with electrode materials, and tolerance to ambient humidity.^28,29 However, its common polymorphic trigonal Li₂ZrCl₆ exhibits intrinsically low ionic conductivity (∼4 × 10⁻⁴ S cm⁻¹ at 30 °C), markedly inferior to those of other superionic conductors by more than an order of magnitude, presenting a critical bottleneck to practical implementation. To overcome this limitation, aliovalent substitution has been widely explored as a viable strategy to enhance the ionic conductivity of Li₂ZrCl₆. In particular, incorporation of trivalent cations^29–33 such as Y³⁺, Er³⁺, Dy³⁺, In³⁺, or Fe³⁺ was proven to be capable of forming stable solid solutions with significantly improved ionic conductivities reaching ∼10⁻³ S cm⁻¹. However, most of these substitution approaches were possible only with the scarce elements that could form a similar trigonal framework, facing fundamental limitations regarding economic viability.^30–33 Substituting them with an Earth-abundant element like Fe³⁺ often led to poor cycling performance due to the undesirable redox activity of Fe.²⁹ Moreover, despite extensive experimental efforts in cation substitution strategies, it remains elusive how a small amount of dopants could enhance the overall lithium diffusion kinetics in the structure, obscuring the choice of rational dopant strategies.

Herein, we investigate the governing factors of lithium-ion transport in the trigonal Li₂ZrCl₆ and reveal general cation interactions that critically shape the energy landscape of diffusion paths in the structure.^24,34 By systematically examining the pristine Li₂ZrCl₆ with a low-valent-dopant substituted system, such as divalent Zn²⁺, we elucidate how the presence of dopants impacts overall and site-specific lithium-ion mobility. It is shown that the effective interaction between dopant and lithium yields a two-fold effect on reshaping the lithium percolation landscape. The partial substitution of Zr⁴⁺ with a low-valent dopant unlocks face-sharing pathways that were previously inaccessible due to the strong electrostatic repulsion from Zr⁴⁺. Moreover, this additional accessibility culminates in local lithium accumulation, which in turn triggers correlated migration behavior. Consequently, the additional accessible pathways and cooperative motion establish a robust and interconnected percolation network for superior ionic transport. Building on this insight, we synthesized low-valent-dopant-substituted systems and verified that a significant enhancement in the ionic conductivity is possible, not only for the Zn-doped case (Li_2+2xZn_xZr_1−xCl₆ (0 < x ≤ 0.25)) but also for vacancy-mediated substitution, achieving an even greater degree of improvement (Li_2+4xZr_1−xCl₆ (0 < x ≤ 0.25)). Finally, we demonstrate the superior electrochemical performance of all-solid-state batteries employing vacancy-doped Li_2.3Zr_0.925Cl₆ paired with LiNi_0.83Co_0.11Mn_0.06O₂ cathodes. At 0.5C, the cell delivers a high specific capacity of 155 mAh g⁻¹, significantly surpassing that of the pristine Li₂ZrCl₆ cell (120 mAh g⁻¹). Furthermore, the enhanced capacity was complemented by long-term stability, maintaining 75% retention after 300 cycles at 0.5C.

Results and discussion

Aliovalent substitution in trigonal halide superionic conductors

Li₂ZrCl₆ exhibits a layered-type structure characterized by a hexagonal close-packed (hcp) anion framework within the P3̄m1 space group, as depicted in Fig. 1a. Zr ions occupy octahedral sites either at 1a or 2d Wyckoff positions, exhibiting a partially disordered arrangement in the neighboring ab-planes. Lithium ions occupy octahedral sites that share their edges with Zr ions along the ab-plane, resulting in a distinctive inverse honeycomb arrangement (ZrLi₆) similar to LiMn₆ in Li₂MnO₃,³⁵ as shown in the lower panel of Fig. 1a. In this hcp anion framework, lithium migration within the ab-plane should proceed via intermediate tetrahedral sites that share a face with the Zr⁴⁺ octahedron, while interlayer migration along the c-axis should occur through direct hopping to neighboring octahedral sites that edge-share with the Zr⁴⁺ octahedron. It indicates that all primary diffusion pathways are intimately linked to the occupancy and distribution of Zr⁴⁺ octahedra and their strong electrostatic repulsion.^24,36,37 In comparison with the structurally analogous halide solid electrolytes such as Li₃YCl₆ and Li₃InCl₆ with trivalent ions, the migration barrier for lithium ions for both intra- and interplane diffusion would be more significantly affected by the presence of the tetravalent Zr⁴⁺ ion. In this context, we performed a detailed lithium diffusion analysis comparing the intrinsic Li₂ZrCl₆ structure with those doped with low-valent cations. We selected Zn²⁺ as a model low-valent cation due to its similar ionic radius (0.74 Å) to Zr⁴⁺ (0.72 Å) for the structural integrity.³⁸ Furthermore, we examined the hypothetical system with the lowest-valent cation, i.e., zero-valent vacancy, by locally removing the Zr⁴⁺ cation (see SI Note 1 for a discussion on the thermodynamic stability of vacancy-substituted structures).

Fig. 1 Effect of aliovalent substitution on the lithium diffusion properties of trigonal Li₂ZrCl₆. (a) Structural representation of Li₂ZrCl₆, with the Zr ion exhibiting partial occupancy at the Wyckoff 1a and 2d sites. The upper panel delineates the overall connectivity and arrangements of Zr 1a and Zr 2d sites. The lower panel depicts the configurations of individual planes, namely Layers 1 and 2. (b) Extrapolated ionic conductivity at 300 K for each component (overall, c-axis, and ab-plane) of Li₂ZrCl₆, Li_2.17Zn_0.08Zr_0.92Cl₆, and Li_2.33Zr_0.92Cl₆, obtained from AIMD simulations conducted between 500 K and 700 K. The overall conductivity is represented in black, that of the ab-plane is in blue, and that of the c-axis is in red. (c–e) Iso-surface (yellow) of lithium probability density P = P_max/1024 at 500 K for three model structures: (c) pristine Li₂ZrCl₆, (d) Zn²⁺-substituted Li_2.17Zn_0.08Zr_0.92Cl₆, and (e) vacancy-substituted Li_2.33Zr_0.92Cl₆, where the vacancy site is denoted as V_Zr. The lower panels illustrate enlarged views of the diffusion pathways near the substituent site, with dashed boxes highlighting the expanded diffusion network induced by substitution.

The lithium diffusion behaviors of these three systems were first probed by conducting ab initio molecular dynamics (AIMD) simulations (see computational details in the Experimental method section of the SI, along with Fig. S2a–c and S3.). Fig. 1b illustrates the overall ionic conductivities estimated at 300 K along with the ab-plane (blue) and c-axis (red) components for respective cases. It reveals that the ionic conduction of the pristine Li₂ZrCl₆ is highly anisotropic, with fast c-axis conduction and negligible ab-plane diffusion. This indicates that despite the layered hcp framework, the sluggish ab-plane diffusion acts as the primary bottleneck, rendering the macroscopic conduction effectively one-dimensional.²⁴ On the other hand, the results show that low-valent cation doping leads to a significant enhancement of the ab-plane lithium conduction kinetics, yielding a more than two-fold increase in the overall ionic conductivity. The ionic conductivity within the ab-plane increases by over four-fold, while that of the c-axis increases by only a factor of ∼1.5, suggesting that the dominant increase in the ab-plane conduction resulted in the enhancement of the total ionic conductivity. It also indicates that the low-valent cation doping is specifically beneficial in expediting ab-plane diffusion, hinting at a unique path-dependent role of the dopant cation in the trigonal Li₂ZrCl₆ structure.

To visualize the local change in diffusion behaviors upon cation substitution, we analyzed the lithium probability density for pristine Li₂ZrCl₆, Zn-doped Li_2.17Zn_0.08Zr_0.92Cl₆, and vacancy-doped Li_2.33Zr_0.92Cl₆, as depicted in Fig. 1c–e, respectively. The yellow regions depict the connectivity of lithium probability density, i.e., diffusion pathway, while the dashed boxes in each figure highlight the most distinct change observed in the pathways upon dopant substitution. Whereas the pathways near unsubstituted Zr⁴⁺ sites remain largely unchanged for both doped cases, those in the vicinity of the substituted sites broaden progressively, which appear to be greater with the zero-valent vacancy dopant. To elucidate this phenomenon, we meticulously analyzed the hopping rates for each ab-plane and c-axis from the AIMD simulations, which are provided in Fig. S4.³⁹ The results indicated that the hopping rates of both the substituted and neighboring layers increased significantly, consistent with the conductivity enhancement trend. On the other hand, non-substituted layers showed minimal changes in hopping rates in any direction. For the substituted layer, the ab-plane hopping rates increased from 8.8 × 10⁹ s⁻¹ in Li₂ZrCl₆ to 1.70 × 10¹⁰ s⁻¹ in Li_2.17Zn_0.08Zr_0.92Cl₆ and further to 2.18 × 10¹⁰ s⁻¹ in Li_2.33Zr_0.92Cl₆, corresponding to an overall increase of approximately 2.5 times. Similarly, the hopping rate in the neighboring layer increased to more than twice its initial value across the same series. In contrast, the c-axis hopping rates displayed a relatively modest rise, from 3.20 × 10¹⁰ s⁻¹ in Li₂ZrCl₆ to 4.66 × 10¹⁰ s⁻¹ in Li_2.33Zr_0.92Cl₆. These results suggest that the enhancement is related to specific local geometry associated with low-valent dopants in the migration pathway, requiring a more detailed, path-by-path analysis to fully elucidate the governing mechanism.

Localized change in lithium diffusion kinetics induced by aliovalent substitution

We carefully examined local diffusion pathways, which proceed particularly through sites sharing an edge or a face with the dopant site (brown octahedron), as illustrated in Fig. 2a. The lithium diffusion pathways within this specific local environment were identified as Path A and Paths B/C, along c-axis and ab-plane directions, respectively. In Path A, a lithium ion migrates through an intermediate octahedral site that shares an edge with the dopant site. Path B represents the in-plane migration on the intralayer involving tetrahedral intermediate sites face-sharing with the dopant site, while Path C is that involving the octahedral intermediate sites face-sharing with the dopant in the neighboring layer. We found that the broadened diffusion pathway in the vicinity of the dopant sites in Fig. 1d and e is attributed to enhanced accessibility through Paths B and C. As shown in Fig. 2b and S5, the substitutions activate entirely new channels that are inaccessible in pristine Li₂ZrCl₆ especially for Paths B and C. While the strong repulsion from the Zr⁴⁺ ion renders the face-sharing intermediate sites in Paths B and C inaccessible in pristine Li₂ZrCl₆,²⁴ the lower-valence substituents, such as Zn²⁺ and vacancies, could mitigate the electrostatic repulsion, which in turn lowers the activation barriers. Notably, the opening of new diffusion pathways could aid in the formation of a percolated diffusion network within the ab-plane by connecting the isolated diffusion clusters, indicated by a red hexagonal ring in Fig. S5a.²⁴

Fig. 2 Pathway-resolved analysis of local lithium diffusion kinetics as a function of aliovalent substitution. (a) Schematic of the three primary local diffusion pathways around a substituent site. Path A represents c-axis diffusion via an edge-sharing octahedron. Paths B and C represent ab-plane diffusion through face-sharing tetrahedral and octahedral sites, respectively. (b) Lithium probability density isosurfaces for each pathway in the pristine, Zn²⁺-substituted, and vacancy-substituted models. The orange and yellow surfaces represent P = P_max/4 and P = P_max/256, respectively (at 500 K). (c) Calculated hopping rates (left) and activation barrier (right) for each specific pathway (Paths A, B, and C) in the three different models: pristine Li₂ZrCl₆ (red), Li_2.17Zn_0.08Zr_0.92Cl₆ (blue), and Li_2.33Zr_0.92Cl₆ (green).

Hopping rates estimated from AIMD simulations quantitatively supported the observed broadening and opening of the diffusion pathways for all relevant diffusion paths (Paths A, B, and C), as presented in the left panel of Fig. 2c.³⁹ A systematic increase in hopping rate was observed as the valence of the substituted metal cations decreased from Zr⁴⁺ to Zn²⁺ and further to vacancies, most notably for Path B. Vacancy substitution led to an increase of nearly two orders of magnitude in the hopping rate compared to pristine Li₂ZrCl₆ in Path B (from 0.147 to 5.42 and 13.8 × 10¹⁰ s⁻¹ for Li₂ZrCl₆, Li_2.17Zn_0.08Zr_0.92Cl₆, and Li_2.33Zr_0.92Cl₆, respectively). The enhanced kinetics could be further confirmed by nudged elastic band (NEB) calculations to determine the activation barrier associated with each diffusion path, as shown in the right panel of Fig. 2c and Fig. S6. For interlayer transport along the c-axis (Path A), a moderate, stepwise reduction in the barrier is observed, from 254 meV for pristine Li₂ZrCl₆ to 197 meV for Zn²⁺ substitution and 171 meV for vacancy substitution. A far more dramatic effect is observed for the ab-plane face-sharing pathways (Paths B and C). These routes, initially prohibited in the pristine Li₂ZrCl₆ with a high activation barrier (>500 meV), become accessible upon substitution. Compared to the pristine framework, vacancy substitution yields the most dramatic drop in activation barriers for Paths B and C (to approximately 145 and 264 meV, respectively), surpassing even the reductions achieved by Zn²⁺ substitution (212 and 273 meV, respectively). The fact that the magnitude of kinetic enhancement exhibits a strong dependence on the valence state of dopant provides compelling evidence that the local electrostatic environment is the dominant governing factor. Given the shorter distance and stronger electrostatic repulsion between the metal cation and lithium ion at face-sharing sites, the ab-plane pathways exhibit heightened sensitivity to the electrostatic environment governed by the cation valence (see SI Note 2 for a discussion about the steric effect on diffusion kinetics).

Another important consequence of low-valent substitution is the presence of additional lithium content to compensate for the charge balance and its spatial redistribution in the structure. According to the analysis of the lithium occupancy of each ab-plane, shown in Fig. S4, and lithium probability density, shown in Fig. 1c–e, we observed that the extra lithium ions predominantly redistribute into the substituted plane and its adjacent layers, with a strong tendency to aggregate near the low-valent dopant sites. Given that the collective motion of lithium ions inherently dictates the macroscopic diffusion within the Li₂ZrCl₆ framework,^40,41 we elucidate that this localized aggregation of extra lithium ions effectively establishes lithium-rich channels conducive to rapid lithium conduction⁴² near substituted planes, amplifying the probability of concerted migration and thereby lowering the overall activation barriers.^43,44 Crucially, van Hove correlation analysis for our materials, a mathematical tool used to describe how the positions of ions are correlated in both space and time, validates this physical picture, confirming that the collective transport mechanism remains dominant and becomes strengthened in our doped system (Fig. S8). Taken together, low-valent substitution not only reduces the activation barrier for lithium diffusion by modulating metal cation–lithium interactions but also provides highly favorable conditions for concerted migration by establishing lithium-rich channels, thereby improving lithium hopping kinetics.

Synthesis and evaluation of the doped Li₂ZrCl₆

Inspired by our theoretical guidance, we synthesized the Zn²⁺- and vacancy-substituted series, Li_2+2xZn_xZr_1−xCl₆ and Li_2+4xZr_1−xCl₆ (0 < x ≤ 0.25) via a mechano-chemical route using stoichiometric mixtures of metal chlorides. Fig. 3a and b present the X-ray diffraction data for the structural analysis of Li_2+2xZn_xZr_1−xCl₆ and Li_2+4xZr_1−xCl₆, respectively, which indicate that both series retain the characteristic trigonal symmetry of Li₂ZrCl₆. The pristine framework remains largely intact at low dopant concentrations up to x = 0.1, confirming the successful incorporation of the dopants into the host lattice without significant structural distortion. Rietveld refinement results for X-ray diffraction (Fig. S9 and S10) further substantiate the structural integrity of the trigonal host across the substitution series. However, the slight transition becomes apparent at a higher substitution level of x = 0.25, with additional reflections alongside the parent peaks, which correspond to those typically seen in disordered cubic close-packed (ccp) halide structures (e.g., LiCl, Li₃InCl₆,¹⁷ and Li₃ScCl₆ ¹⁹). This tendency of partial transition aligns with established principles that the ccp structure is preferred over the hcp framework with higher ionic potentials of lithium and metal in layered lithium halides⁴⁵ (see SI Note 3 for details). We further scrutinized the detailed structures of representative samples, Li₂ZrCl₆, Li_2.2Zn_0.1Zr_0.9Cl₆, and Li_2.3Zr_0.925Cl₆, by performing high-resolution neutron diffraction, as presented in Fig. 3c–e. The Rietveld refinements confirmed the trigonal hcp lattice as a dominant phase with the successful incorporation of the dopants into the host, particularly within the doping range of x ≤ 0.1 (see Tables S1–3 for details). Additionally, we conducted X-ray absorption spectroscopy (XAS) analysis on the doped samples, which did not reveal any significant structural deviations, except for the slight atomic rearrangement expected for dopant substitution (Fig. S11 and S12 and Table S4).

Fig. 3 Structural characterization and ionic transport properties of Li_2+2xZn_xZr_1−xCl₆ and Li_2+4xZr_1−xCl₆ (x = 0–0.25). (a and b) X-ray diffraction patterns for (a) the Li_2+2xZn_xZr_1−xCl₆ series and (b) the Li_2+4xZr_1−xCl₆ series (x = 0–0.25). Reference patterns for pristine Li₂ZrCl₆ (hcp) and LiCl (ccp) are provided for structural comparison. (c–e) Rietveld refinement profiles of neutron diffraction data for (c) Li₂ZrCl₆, (d) Li_2.2Zn_0.1Zr_0.9Cl₆, and (e) Li_2.3Zr_0.925Cl₆. (f and g) Arrhenius plots showing the ionic conductivities (left axis, blue) and the corresponding activation energies (right axis, orange) for (f) the Zn²⁺-substituted series and (g) the vacancy-substituted series.

The ionic transport properties of these samples were comparatively analyzed and summarized with respect to the measured ionic conductivity and activation barrier in Fig. 3f and g and S13. The vacancy-substituted series exhibits a marked enhancement, reaching a peak conductivity of 8 × 10⁻⁴ S cm⁻¹ at 30 °C for Li_2.3Zr_0.925Cl₆ (x = 0.075), which is more than twice that of pristine Li₂ZrCl₆ (3.6 × 10⁻⁴ S cm⁻¹ at 30 °C). The Zn²⁺-substituted series also shows an increase in conductivity, with a maximum value of 6.4 × 10⁻⁴ S cm⁻¹ for Li_2.2Zn_0.1Zr_0.9Cl₆ (x = 0.1) at 30 °C. A clear performance hierarchy emerges as the effective charge of the cation sublattice is systematically reduced, whereby the ionic conductivity exhibits a commensurate increase as the valency decreases from Zr⁴⁺ to Zn²⁺ and ultimately to zero-valent vacancies. We further note that the relative densities of the cold-pressed pellets are comparable across all three compositions (75.8%, 77.8%, and 76.8% for Li₂ZrCl₆, Li_2.2Zn_0.1Zr_0.9Cl₆, and Li_2.3Zr_0.925Cl₆, respectively), confirming that the observed enhancements are intrinsic to the doping effect rather than the densification degree (Fig. S14). Both Li_2.3Zr_0.925Cl₆ and Li_2.2Zn_0.1Zr_0.9Cl₆ displayed lower activation energies (0.351 eV and 0.355 eV, respectively) in comparison with Li₂ZrCl₆ (0.370 eV), strongly suggesting that a weakened electrostatic environment reshapes the energy landscape for lithium-ion migration. Notably, the ionic conductivity of both series follows a volcano-shaped dependence on the substitution level, where the conductivity decline beyond the optimal doping level is attributed to the partial phase transition to the disordered ccp phase with intrinsically lower mobility^46–49 (see SI Notes 3 for details). Nevertheless, the realization of significant conductivity enhancement with only a small amount of substitution, even with such structural constraints, strongly validates the efficacy of our theoretical design.

Beyond the improved ionic transport, it is crucial to verify whether these doping strategies compromise the intrinsic electrochemical stability of the host framework. To further assess this, cyclic voltammetry was performed on the pristine Li₂ZrCl₆ alongside the optimal compositions of each doping strategy, namely Li_2.2Zn_0.1Zr_0.9Cl₆ and Li_2.3Zr_0.925Cl₆ (Fig. S15 and S16). The electrochemical stability window is largely preserved upon doping, with no significant shift in either the oxidative or reductive onset potential (∼4.0 V and ∼1.8 V vs. Li/Li⁺, respectively). In particular, the oxidative current is predominantly confined to the first cycle and diminishes markedly in subsequent sweeps, indicative of stable passivation layer formation at high voltages.

Electrochemical performance of the doped-Li₂ZrCl₆ in solid-state batteries

The electrochemical properties of the solid electrolytes were tested for samples with the highest ionic conductivities, i.e., Li_2.2Zn_0.1Zr_0.9Cl₆ and Li_2.3Zr_0.925Cl₆, in a solid-state battery setup using a commercial-level single crystal high-nickel cathode (LiNi_0.83Co_0.11Mn_0.06O₂) as the active material. Fig. 4a presents the first cycle profile of these cells at 0.1C (where 1C = 200 mA g⁻¹) within a voltage range of 3.0–4.3 V (vs. Li/Li⁺). At 0.1C, the solid-state cells employing doped electrolytes exhibited electrochemical profiles comparable to the pristine Li₂ZrCl₆. They delivered comparably high initial discharge capacities of 194.5 and 196 mAh g⁻¹ with initial coulombic efficiencies of 86.6% and 88%, respectively, mirroring the robust performance of its pristine counterpart (197 mAh g⁻¹, 86%). This confirms that the electrochemical reversibility inherent to the halide framework is well-preserved even after doping. On the other hand, the rate capability of the cells shows a strong correlation with the ionic conductivity of the solid electrolyte, with performance disparities becoming more pronounced at higher current densities, as shown in Fig. 4b. Notably, at a practical rate of 0.5C, the cell utilizing Li_2.3Zr_0.925Cl₆ outperformed its counterparts, delivering a discharge capacity of 155 mAh g⁻¹, which corresponds to a retention of 78.9% relative to the specific capacity at 0.1C. This stands in clear distinction to the Li_2.2Zn_0.1Zr_0.9Cl₆ and pristine Li₂ZrCl₆ cells, which yielded lower capacities of 137 mAh g⁻¹ and 120 mAh g⁻¹ with corresponding retentions of 70.7% and 60.8%, respectively. The mechanistic origin of this kinetic hierarchy was further elucidated by the galvanostatic intermittent titration technique and electrochemical impedance spectroscopy analyses (Fig. S17 and S18). The progressive substitution from Zr⁴⁺ to Zn²⁺ and to zero-valent vacancies systematically reduces both the bulk electrolyte resistance and the interfacial charge transfer resistance, indicating that the enhanced ionic conductivity facilitates Li⁺ transport at both the bulk and interfacial levels within the composite cathode. Beyond its superior rate capability, the cell incorporating Li_2.3Zr_0.925Cl₆ demonstrated robust long-term durability. It successfully delivered a higher discharge capacity at 0.5C without compromising the cyclability inherent to the pristine framework, as shown in Fig. 4c. After 300 cycles, the cell retained approximately 75% of its initial capacity, a value comparable to those of its pristine counterparts. This confirms that significant kinetic enhancement was achieved without sacrificing electrochemical stability. The versatility of the optimized electrolyte was further corroborated in LiCoO₂ cells, as illustrated in Fig. S19 and S20. Specifically, as highlighted in Fig. S20, the cell utilizing Li_2.3Zr_0.925Cl₆ could also retain the high-nickel cathode performance, demonstrating outstanding rate capability and durability. At 1C, it delivered a high specific capacity of 118 mAh g⁻¹ with 84% retention relative to 0.1C, outperforming Li_2.2Zn_0.1Zr_0.9Cl₆ (113 mAh g⁻¹, 81%) and the pristine Li₂ZrCl₆ (106 mAh g⁻¹, 75%). Furthermore, superior long-term stability was achieved, maintaining 75% of its capacity after 500 cycles compared to 71% for the pristine cell under identical conditions.

Fig. 4 Electrochemical performance and interfacial stability of all-solid-state batteries employing pristine and substituted Li₂ZrCl₆ solid electrolytes. (a) Initial charge/discharge voltage profiles at 0.1C for all-solid-state cells employing pristine Li₂ZrCl₆, Zn-doped Li_2.2Zn_0.1Zr_0.9Cl₆, and vacancy-doped Li_2.3Zr_0.925Cl₆ as the solid electrolyte with a single-crystal LiNi_0.83Co_0.11Mn_0.06O₂ cathode. (b) Rate capability comparison of the corresponding cells at current densities ranging from 0.1C to 1C. (c) Cycling performance of the corresponding cells after a formation cycle at a rate of 0.5C over 300 cycles. Filled and open circles represent specific capacity (left axis) and coulombic efficiency (right axis), respectively. (d) Time-of-flight secondary ion mass spectrometry results for the composite cathode before and after 300 cycles. The upper panels display signals corresponding to cathode-derived fragments (NiO⁻, MnO⁻ and CoO⁻), while the lower panels show electrolyte-derived fragments (ClO⁻, ZrO⁻ and ZrO₂⁻).

To verify the chemical stability of the halide solid electrolyte and its compatibility with the oxide cathode, time-of-flight secondary ion mass spectrometry (ToF-SIMS) was employed to probe the elemental distribution within cycled composite electrodes consisting of LiNi_0.83Co_0.11Mn_0.06O₂ and pristine-, Zn-, or vacancy-substituted Li₂ZrCl₆. As illustrated in Fig. 4d, the distribution of characteristic secondary ion fragments corresponding to the oxide cathode lattice (i.e., Ni–O, Co–O, and Mn–O) remained largely consistent after cycling across all evaluated compositions, indicating that the chemical integrity of the active material is preserved without significant degradation at the interface. Simultaneously, we monitored the emergence of Cl–O and Zr–O signals, which are typical diagnostic indicators of interfacial side reactions or electrolyte decomposition.⁵⁰ The insignificant accumulation of these species confirms the superior electrochemical stability of our system, suggesting that no substantial parasitic reactions occur during operation. Furthermore, this chemical robustness was consistently observed in LiCoO₂-based cells. As shown in Fig. S21 and S22, both XPS and ToF-SIMS analyses revealed minor signal deviations even after extensive cycling, confirming the absence of substantial interfacial degradation and the electrochemical stability of our system.

Summary

This work provides a comprehensive, atomistic-level understanding of the mechanism by which aliovalent substitution enhances lithium-ion transport in the trigonal Li₂ZrCl₆ framework, establishing a principle for the rational design of cost-effective halide electrolytes. Our theoretical investigation revealed that reducing the electrostatic repulsive force exerted by the metal cation is a critical factor in enhancing the lithium diffusion kinetics of aliovalent substitution in the Li₂ZrCl₆ structure. Substituting Zr⁴⁺ with a less repulsive ion not only promotes lithium diffusion near the substituted site with lithium aggregation, facilitating concerted diffusion behavior, but also opens up new diffusion pathways, which have been prohibited due to strong electrostatic repulsion from Zr⁴⁺ in pristine Li₂ZrCl₆. As a proof-of-concept for our theoretical findings, we synthesized both Zn²⁺- and vacancy-substituted series (Li_2+2xZn_xZr_1−xCl₆ and Li_2+4xZr_1−xCl₆, where x = 0–0.25). The vacancy-doped Li_2.3Zr_0.925Cl₆ exhibited the highest ionic conductivity of 0.8 mS cm⁻¹, surpassing that of the pristine Li₂ZrCl₆ and best-performing Zn²⁺-substituted compound, thereby corroborating computational predictions regarding the performance hierarchy. Furthermore, the optimized composition, Li_2.3Zr_0.925Cl₆, demonstrated superior rate capability and robust electrochemical stability, with ex situ analysis confirming negligible interfacial degradation. By introducing ‘zero-cost’ metal vacancies as dopants into the cost-effective Li₂ZrCl₆ framework, our strategy offers an economically viable route to high-voltage halide superionic conductors, bypassing the reliance on expensive rare-earth metals (e.g., Y, In, or Sc).

Author contributions

J. N. and S. Y. contributed equally to this work. Joohyeon Noh: conceptualization, data curation, formal analysis, investigation, methodology, validation, visualization, writing – original draft, and writing – review and editing. Seungju Yu: data curation, formal analysis, methodology, validation, visualization, writing – original draft, and writing – review and editing. Sunyoung Lee: investigation, methodology, and validation. Wonju Kim: methodology and validation. Kyungho Yoon: methodology and validation. Sangwook Han: methodology. Junhyuk Song: methodology. Kyungbae Oh: methodology and formal analysis. Chanwoong Park: validation. Daero Won: methodology. Geunji Choi: methodology. Kangtaek Lee: methodology. Kisuk Kang: conceptualization, investigation, methodology, project administration, validation, visualization, supervision, writing – original draft, and writing – review and editing.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. The SI includes the experimental methods and computational details, supplementary notes on thermodynamic stability and phase analysis, and supplementary figures and tables containing the computational (diffusion and energy) data, structural characterization (X-ray and neutron diffraction and X-ray absorption spectroscopy), and electrochemical measurement data. See DOI: https://doi.org/10.1039/d6eb00070c.

Acknowledgements

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science and ICT, No. RS-2023-00261543 and RS-2021-NR057375). This work was also supported by the Research Program of LOTTE Energy Materials.

References

1. Q. Zhao, S. Stalin, C.-Z. Zhao and L. A. Archer, Nat. Rev. Mater., 2020, 5, 229–252.
2. J. Janek and W. G. Zeier, Nat. Energy, 2016, 1, 16141.
3. C. Wang and X. Sun, Engineering, 2023, 21, 32–35.
4. J. T. Frith, M. J. Lacey and U. Ulissi, Nat. Commun., 2023, 14, 420.
5. Z. Zhang, Y. Shao, B. Lotsch, Y.-S. Hu, H. Li, J. Janek, L. F. Nazar, C.-W. Nan, J. Maier, M. Armand and L. Chen, Energy Environ. Sci., 2018, 11, 1945–1976.
6. R. Chen, Q. Li, X. Yu, L. Chen and H. Li, Chem. Rev., 2020, 120, 6820–6877.
7. J. Janek and W. G. Zeier, Nat. Energy, 2023, 8, 230–240.
8. E. P. Alsaç, D. L. Nelson, S. G. Yoon, K. A. Cavallaro, C. Wang, S. E. Sandoval, U. D. Eze, W. J. Jeong and M. T. McDowell, Chem. Rev., 2025, 125, 2009–2119.
9. X. Li, J. Liang, X. Yang, K. R. Adair, C. Wang, F. Zhao and X. Sun, Energy Environ. Sci., 2020, 13, 1429–1461.
10. C. Wang, J. Liang, J. T. Kim and X. Sun, Sci. Adv., 2022, 8, eadc9516.
11. Y. Ren, T. Danner, A. Moy, M. Finsterbusch, T. Hamann, J. Dippell, T. Fuchs, M. Müller, R. Hoft, A. Weber, L. A. Curtiss, P. Zapol, M. Klenk, A. T. Ngo, P. Barai, B. C. Wood, R. Shi, L. F. Wan, T. W. Heo, M. Engels, J. Nanda, F. H. Richter, A. Latz, V. Srinivasan, J. Janek, J. Sakamoto, E. D. Wachsman and D. Fattakhova-Rohlfing, Adv. Energy Mater., 2022, 13, 2201939.
12. S.-K. Jung, H. Gwon, S.-S. Lee, H. Kim, J. C. Lee, J. G. Chung, S. Y. Park, Y. Aihara and D. Im, J. Mater. Chem. A, 2019, 7, 22967–22976.
13. K. Yoon, J.-J. Kim, W. M. Seong, M. H. Lee and K. Kang, Sci. Rep., 2018, 8, 8066.
14. K. Yoon, H. Kim, S. Han, T.-S. Chan, K.-H. Ko, S. Jo, J. Park, S. Kim, S. Lee, J. Noh, W. Kim, J. Lim and K. Kang, Appl. Phys. Rev., 2022, 9, 031403.
15. Y. Zhai, W. Kang, Y. Li, N. Zhang, X. Liu, R. An, C. Yang, X. Zhu, Q. Huang, L. Chen, D. Cao, M. Wang, Y. Lu, J. Li, Z. Xiong, C. Feng, H. Jin, Y. Guan, Y. Su, F. Wu and N. Li, Energy Mater. Adv., 2025, 6, 0163.
16. T. Asano, A. Sakai, S. Ouchi, M. Sakaida, A. Miyazaki and S. Hasegawa, Adv. Mater., 2018, 30, e1803075.
17. X. Li, J. Liang, J. Luo, M. Norouzi Banis, C. Wang, W. Li, S. Deng, C. Yu, F. Zhao, Y. Hu, T.-K. Sham, L. Zhang, S. Zhao, S. Lu, H. Huang, R. Li, K. R. Adair and X. Sun, Energy Environ. Sci., 2019, 12, 2665–2671.
18. L. Zhou, C. Y. Kwok, A. Shyamsunder, Q. Zhang, X. Wu and L. F. Nazar, Energy Environ. Sci., 2020, 13, 2056–2063.
19. J. Liang, X. Li, S. Wang, K. R. Adair, W. Li, Y. Zhao, C. Wang, Y. Hu, L. Zhang, S. Zhao, S. Lu, H. Huang, R. Li, Y. Mo and X. Sun, J. Am. Chem. Soc., 2020, 142, 7012–7022.
20. S. Muy, J. Voss, R. Schlem, R. Koerver, S. J. Sedlmaier, F. Maglia, P. Lamp, W. G. Zeier and Y. Shao-Horn, iScience, 2019, 16, 270–282.
21. S. Y. Kim, K. Kaup, K.-H. Park, A. Assoud, L. Zhou, J. Liu, X. Wu and L. F. Nazar, ACS Mater. Lett., 2021, 3, 930–938.
22. J. Liang, E. van der Maas, J. Luo, X. Li, N. Chen, K. R. Adair, W. Li, J. Li, Y. Hu, J. Liu, L. Zhang, S. Zhao, S. Lu, J. Wang, H. Huang, W. Zhao, S. Parnell, R. I. Smith, S. Ganapathy, M. Wagemaker and X. Sun, Adv. Energy Mater., 2022, 12, 2103921.
23. A. Zheng, L. Luo, L. Li, Z. Jiang, S. Ma and J. Yu, J. Mater. Chem. A, 2025, 13, 6067–6074.
24. S. Yu, J. Noh, B. Kim, J.-H. Song, K. Oh, J. Yoo, S. Lee, S.-O. Park, W. Kim, B. Kang, D. Kil and K. Kang, Science, 2023, 382, 573–579.
25. L. Huang, L. Zhang, J. Bi, T. Liu, Y. Zhang, C. Liu, J. Cui, Y. Su, B. Wu and F. Wu, Energy Mater. Adv., 2024, 5, 0092.
26. Z. Liu, P.-H. Chien, S. Wang, S. Song, M. Lu, S. Chen, S. Xia, J. Liu, Y. Mo and H. Chen, Nat. Chem., 2024, 16, 1584–1591.
27. H. Kwak, S. Wang, J. Park, Y. Liu, K. T. Kim, Y. Choi, Y. Mo and Y. S. Jung, ACS Energy Lett., 2022, 7, 1776–1805.
28. K. Wang, Q. Ren, Z. Gu, C. Duan, J. Wang, F. Zhu, Y. Fu, J. Hao, J. Zhu, L. He, C. W. Wang, Y. Lu, J. Ma and C. Ma, Nat. Commun., 2021, 12, 4410.
29. H. Kwak, D. Han, J. Lyoo, J. Park, S. H. Jung, Y. Han, G. Kwon, H. Kim, S. T. Hong, K. W. Nam and Y. S. Jung, Adv. Energy Mater., 2021, 11, 2003190.
30. S. Chen, C. Yu, S. Chen, L. Peng, C. Liao, C. Wei, Z. Wu, S. Cheng and J. Xie, Chin. Chem. Lett., 2022, 33, 4635–4639.
31. Q. Shao, C. Yan, M. Gao, W. Du, J. Chen, Y. Yang, J. Gan, Z. Wu, W. Sun, Y. Jiang, Y. Liu, M. Gao and H. Pan, ACS Appl. Mater. Interfaces, 2022, 14, 8095–8105.
32. S. Chen, C. Yu, C. Wei, Z. Jiang, Z. Zhang, L. Peng, S. Cheng and J. Xie, Energy Mater. Adv., 2023, 4, 0019.
33. Z. Yao, Q. Jia, J. Xiang, F. Tu, J. Shi, Y. Zhang, J. Huang, Y. Jiang, W. Gong and D. Yang, J. Electron. Mater., 2024, 54, 1021–1028.
34. K. Kim, D. Park, H. G. Jung, K. Y. Chung, J. H. Shim, B. C. Wood and S. Yu, Chem. Mater., 2021, 33, 3669–3677.
35. D. Luo, H. Zhu, Y. Xia, Z. Yin, Y. Qin, T. Li, Q. Zhang, L. Gu, Y. Peng, J. Zhang, K. M. Wiaderek, Y. Huang, T. Yang, Y. Tang, S. Lan, Y. Ren, W. Lu, C. M. Wolverton and Q. Liu, Nat. Energy, 2023, 8, 1078–1087.
36. K. S. Kang, Y. S. Meng, J. Breger, C. P. Grey and G. Ceder, Science, 2006, 311, 977–980.
37. A. Van der Ven and G. Ceder, J. Power Sources, 2001, 97–98, 529–531.
38. R. D. Shannon, Acta Crystallogr., Sect. A, 1976, 32, 751–767.
39. N. J. J. de Klerk, E. van der Maas and M. Wagemaker, ACS Appl. Energy Mater., 2018, 1, 3230–3242.
40. H. Guo, V. Koverga, S. C. Selvaraj and A. T. Ngo, ACS Appl. Energy Mater., 2025, 8, 14773–14790.
41. Y. C. Lee and S. C. Jung, Electrochim. Acta, 2024, 497, 144632.
42. G. Yang, X. Liang, S. Zheng, H. Chen, W. Zhang, S. Li and F. Pan, eScience, 2022, 2, 79–86.
43. X. He, Y. Zhu and Y. Mo, Nat. Commun., 2017, 8, 15893.
44. Y. Xiao, K. Jun, Y. Wang, L. J. Miara, Q. Tu and G. Ceder, Adv. Energy Mater., 2021, 11, 2101437.
45. Q. Wang, Y. Zhou, X. Wang, H. Guo, S. Gong, Z. Yao, F. Wu, J. Wang, S. Ganapathy, X. Bai, B. Li, C. Zhao, J. Janek and M. Wagemaker, Nat. Commun., 2024, 15, 1050.
46. J. Noh, M. Kim, S. Yu, W. Kim, J. Yoo, S. Lee, S. Han, D. Won, C. Park, G. Choi and K. Kang, Solid State Ionics, 2025, 428, 116920.
47. H.-J. Jeon, Y. Subramanian and K.-S. Ryu, J. Power Sources, 2024, 602, 234343.
48. X. Liu, F. Mi and C. Sun, Chem. Commun., 2025, 61, 1144–1147.
49. H. Zhang, Z. Yu, H. Chen, Y. Zhou, X. Huang and B. Tian, J. Energy Chem., 2023, 79, 348–356.
50. I. Kochetkov, T.-T. Zuo, R. Ruess, B. Singh, L. Zhou, K. Kaup, J. Janek and L. Nazar, Energy Environ. Sci., 2022, 15, 3933–3944.

---

Footnote

† These authors contributed equally to this paper.

---