A Li-enriched amorphous Zr-based oxychloride solid electrolyte for high-rate and long-cycling all-solid-state ultrahigh-nickel cathodes

Abstract

The Zr-based chloride solid electrolyte (SE) Li₂ZrCl₆ is well compatible with 4 V-class cathodes and cost-effective, yet its low conductivity (<1 mS cm⁻¹) would restrain the capacity delivery of the electrodes especially at high rates. Herein, we further elevated the room-temperature ionic conductivity of Zr-based lithium oxychloride to 2.11 mS cm⁻¹ in Li_2.15Zr_0.85In_0.15Cl₄O by tuning its Li content through partial substitution of Zr⁴⁺ with In³⁺ in the formula Li_2+xZr_1−xIn_xCl₄O. Cold-pressed Li_2.15Zr_0.85In_0.15Cl₄O presents both a dense morphology arising from its amorphous phase and enhanced ionic conductivity, which is significantly higher than that of Li₂ZrCl₆, facilitating better solid–solid contact and improved reaction kinetics in the composite cathodes. As a result, the all-solid-state cathode coupling Li_2.15Zr_0.85In_0.15Cl₄O and LiNi_0.92Co_0.03Mn_0.05O₂ shows a capacity of 175.3 mAh g⁻¹ at 4C and a retention of 91.44% for 450 cycles when charged to 4.3 V vs. Li⁺/Li. More attractively, as the charge upper limit increases to 4.8 V, Li_2.15Zr_0.85In_0.15Cl₄O also enables ultrahigh-nickel cathodes to cycle for over 250 cycles with a capacity retention of 81.86%.

---

Jipeng Fu

Dr Jipeng Fu received his PhD from Changchun Institute of Applied Chemistry, Chinese Academy of Sciences (supervised by Academician Hongjie Zhang) and Technische Universität Dresden, Germany (supervised by Prof. Henning Heuer). His research focuses on advanced materials for sustainable energy applications, including solid-state batteries and optoelectronic detection instruments. He has held research positions at the Fraunhofer Institute (Germany) and the Chinese Academy of Sciences, and he was a visiting scholar at ETH Zurich (Switzerland) and the University of Technology Sydney (Australia). Dr Fu has authored the textbook Semiconductor Physics and Devices and co-authored the Wiley-VCH Book Energy Storage Materials Characterization. He has been honored as a Lindau Nobel Laureate Young Scientist and Journal of Materials Chemistry Emerging Investigator.

1 Introduction

The pursuit of higher energy density remains an enduring challenge in power lithium batteries, so the evolution of cathode materials from lithium iron phosphate (LiFePO₄) to ultrahigh-nickel ternary layered oxides (LiNi_xCo_yMn_1−x−yO₂, x > 0.9) with higher voltages (>4 V) and larger capacities (>200 mAh g⁻¹) is an anticipated trend. However, the high reactivity of ultrahigh-nickel oxide cathodes inevitably aggravates the interfacial instability and severe safety risk in commercial lithium-ion batteries, particularly when paired with highly reactive and flammable organic liquid electrolytes.

All-solid-state lithium batteries (ASSLBs), which employ non-flammable solid electrolytes (SEs) in lieu of liquid electrolytes, have emerged as highly promising contenders with improved energy density and safety. Considering the rigid nature and high surface activity of ultrahigh-nickel cathodes, the selection of a suitable SE with sufficient cathode compatibility, including good electrolyte/cathode solid–solid contact and sufficient oxidation resistance, is critical for ensuring stable, long-term cycling. Among inorganic SEs, oxides (e.g., Li₇La₃Zr₂O₁₂) are renowned for their relatively high chemical stability, yet their high rigidity leads to large interfacial contact resistance and necessitates extra interfacial wetting modifications when paired with ultrahigh-nickel cathodes.^1–4 In comparison, sulfide SEs (e.g., Li₆PS₅Cl) exhibit excellent ionic conductivity and sufficient deformability,^5–7 but suffer from rapid interfacial degradation and potential combustion due to their narrow electrochemical window and the release of flammable sulfide species (e.g., H₂S) during cycling.⁸ On the other hand, chloride SEs integrate the advantages of a non-flammable nature and excellent compatibility with ultrahigh-nickel cathodes, enabling safe and stable cycling.^9,10 Nevertheless, the reliance on rare metals (e.g., Sc and Ta) somewhat raises concerns regarding their cost and long-term sustainability.

Among various metal chloride SEs, Li₂ZrCl₆ has garnered substantial attention due to its notable cost-effectiveness and good compatibility with 4 V-class cathodes.^11,12 Nevertheless, its relatively low room-temperature ionic conductivity,^11,12 typically below 1 mS cm⁻¹, still hinders the capacity delivery of ASSLBs using Li₂ZrCl₆ as the catholyte especially at high rates (>1C). To address this issue, partial substitution of chloride with oxygen has been proven effective to obtain an elevated ionic conductivity of up to 1.35 mS cm⁻¹.^13,14 Moreover, compared to ball-milled Li₂ZrCl₆ composed of both an amorphous matrix and P3m1 crystalline phase,¹² Zr-based oxychlorides with a higher amorphous degree can further mitigate interface issues arising from crystalline boundaries. Hence, compared to its chloride counterpart (Li₂ZrCl₆), oxygen substitution endows amorphous Zr-based oxychloride SEs with enhanced kinetics and better solid–solid contact in ASSLBs (Fig. 1a).

Fig. 1 Morphologies and ionic conductivities of Li₂ZrCl₆ and Li₂ZrCl₄O SEs. (a) Schematic of the ASSLBs assembled with Li₂ZrCl₆ and Li₂ZrCl₄O SEs. (b and c) SEM images of the surfaces of cold-pressed Li₂ZrCl₆ SE. (d and e) SEM images of the surfaces of cold-pressed Li₂ZrCl₄O SE. (f) Nyquist plots from electrochemical impedance spectroscopy (EIS) test of Li₂ZrCl₆ and Li₂ZrCl₄O SEs. (g) Ionic conductivities of Li₂ZrCl₆ and Li₂ZrCl₄O SEs at 25 °C. (h) Temperature (T)–dependent ionic conductivities (σ) of the as-prepared Li₂ZrCl₆ and Li₂ZrCl₄O SEs. (i) Activation energies (E_a) of Li₂ZrCl₆ and Li₂ZrCl₄O SEs.

Herein, we develop an amorphous Li_2.15Zr_0.85In_0.15Cl₄O SE with elevated lithium-ion conductivity of up to 2.11 mS cm⁻¹ at 25 °C and good compatibility with the ultrahigh-nickel LiNi_0.92Co_0.03Mn_0.05O₂ (NCM92) cathode. Besides the higher ionic conductivity compared with Li₂ZrCl₆, Li_2.15Zr_0.85In_0.15Cl₄O also presents better deformability, benefiting from its higher amorphous degree, promising better solid–solid contact when paired with NCM92 in the composite cathode. Moreover, it demonstrates good reversibility, maintaining a capacity retention of 91.44% after 450 cycles at a cut-off voltage of 4.3 V vs. Li⁺/Li. Notably, even when the cut-off voltage is elevated to 4.8 V vs. Li⁺/Li, the ASSLB can still operate stably for 250 cycles, with a capacity retention of over 81.86%. These findings highlight the robustness and adaptability of the proposed SEs under more demanding operating conditions, paving the way for the development of ASSLBs based on amorphous SEs with a superior rate performance and extended cycle life.

2 Methods

Material synthesis

The raw materials for synthesizing the (oxy)chloride SEs Li₂ZrCl₆, Li₂ZrCl₄O, and Li_2+xZr_1−xIn_xCl₄O (x = 0.05, 0.10, 0.15, 0.20, 0.25, and 0.30) were stoichiometric mixtures of ZrCl₄ (Energy Chemical & 3A, 99.9%), Li₂O (Energy Chemical & 3A, 99%), InCl₃ (Aladdin, 99.99%), and LiCl (Aladdin, 99.99%). The synthesis procedure involved an initial low-speed ball milling step at 100 rpm for 2 h to ensure homogeneous mixing, followed by high-speed ball milling at 850 rpm for 22 h. Li_5.5PS_4.5Cl_0.8Br_0.7 was prepared from a stoichiometric mixture of Li₂S (Aladdin, 99.5%), LiBr (Energy Chemical & 3A, 99%), LiCl (Aladdin, 99.99%), and P₂S₅ (Aladdin, 98.5%). The synthesis procedure involved an initial low-speed ball milling step at 400 rpm for 10 h to ensure homogenization, followed by annealing at 450 °C for 8 h. All procedures were conducted in an argon-filled glovebox (H₂O, O₂ < 0.1 ppm). The SE products were then stored in the glovebox for further use.

Characterization methods

Lab-based X-ray diffraction (XRD) measurements were conducted using a Bruker diffractometer with Cu Kα radiation (λ = 1.5406 Å). To prevent air exposure, Kapton tape was used to cover the sample holder. X-ray photoelectron spectroscopy (XPS) was performed using a Thermo Fisher Scientific K-Alpha instrument with a monochromatic Al Kα X-ray source. Zr K-edge X-ray absorption fine structure spectra (XAFS) data were collected using a RapidXAFS 2M instrument from Anhui Absorption Spectroscopy Analysis Instrument Co., Ltd. The spectra were recorded in transmission mode under operating conditions of 20 kV and 20 mA. The XAFS data were processed using the ATHENA module implemented in the IFEFFIT software package,¹⁵ and least-square curve-fitting was accomplished by using the ARTEMIS module of IFEFFIT.¹⁶ SEM images were obtained using a JEOL-JSM-6700F scanning electron microscope with an acceleration voltage of 10 kV. The depth morphologies of the composite cathode were captured with the application of a Ga⁺ focused ion beam (FIB) at an accelerated voltage of 20 kV and current of 2.5 nA. ⁷Li 1D magic-angle spinning (MAS) NMR measurements were performed on a Bruker AVANCE NEO 600 WB spectrometer (B₀ = 14.1 T, Larmor frequency ω₀ = 233.24 MHz for ⁷Li) with a 3.2 mm HXY MAS probe at room temperature. To avoid deterioration of the SEs under air, all powder samples were sealed in ZrO₂ tubes with VESPEL caps in an Ar-filled glovebox (H₂O, O₂ < 0.1 ppm) before measurement.¹⁷ MAS was applied to acquire highly resolved spectra, with a spinning rate of 22 kHz for 1D NMR experiments. All ⁷Li chemical shifts were calibrated with 1 M LiCl (aq.).¹⁸

Measurements of ionic conductivities

The ionic conductivity of the (oxy)chloride SEs was evaluated using electrochemical impedance spectroscopy (EIS). Approximately 120 mg of SE powder was cold-pressed into pellets under 4.0 tons pressure and sandwiched between two stainless steel rods. The thickness of the pellets ranged from 0.03 to 0.08 cm. The ionic conductivity test was performed using an electrochemical workstation (EC-Lab Electrochemistry SP300, Biologic) at a frequency in the range of 7 MHz to 10 Hz.

Linear sweep voltammetry (LSV) test

For linear sweep voltammetry (LSV) measurements, 80 mg of (oxy)chloride SE (Li₂ZrCl₆, Li₂ZrCl₄O or Li_2.15Zr_0.85In_0.15Cl₄O) was filled in a zirconia (ZrO₂) cylinder (10 mm diameter) and pressed at 2 tons for 30 s, followed by the addition of 30 mg Li_5.5PS_4.5Cl_0.8Br_0.7 to the cylinder and pressed at 3 tons for another 30 s. To prepare the (oxy)chloride SE-carbon composite, the (oxy)chloride SE and vapor-grown carbon fiber (VGCF) were mixed in a weight ratio of 7 : 3 and hand-ground in an agate mortar for 15 min. 5 mg of the composite was put onto the (oxy)chloride SE side to serve as the working electrode and pressed at 4 tons for another 1 min. On the Li_5.5PS_4.5Cl_0.8Br_0.7 side of the SE pellet, a thin lithium (Li) foil (Tianjin Zhongneng Lithium Industry Co., Ltd) with a thickness of about 0.2 mm was attached. The cell was then placed into a stainless-steel casing with a constant applied pressure of about 1.0 tons. The LSV measurement was performed at a scan rate of 1.0 mV s⁻¹.

Assembly and electrochemical characterization of ASSLBs

ASSLBs were fabricated by integrating Li₂ZrCl₆ or Li_2.15Zr_0.85In_0.15Cl₄O SEs with a LiNi_0.92Co_0.03Mn_0.05O₂ (NCM92) cathode, Li_5.5PS_4.5Cl_0.8Br_0.7 and an In–Li alloy anode. For the preparation of the cathode composite, NCM92 was blended with Li₂ZrCl₄O or Li_2.15Zr_0.85In_0.15Cl₄O SEs, polytetrafluoroethylene (PTFE) binder, and VGCF conductive additive at a mass ratio of 60 : 35 : 3 : 2. Initially, the mixtures were manually ground in an agate mortar for 10 min. Subsequently, they were further mixed using a miniature vibration mixer (MSK-SFM-12 M, Hefei Kejing Materials Technology Co., Ltd) for an additional 10 min to obtain the final cathode composite powders. In the assembly process, approximately 80 mg of the SE powder was first placed in a ZrO₂ cylinder with a diameter of 10 mm, and then pressed at a pressure of 1.0 ton for 10 s to form an (oxy)chloride SE layer. Subsequently, 2–3 mg of cathode composite powder was evenly spread over one side of the (oxy)chloride SE layer, and then pressed at 2 tons for 10 s. To prevent side-reactions between the (oxy)chloride SE and anode, approximately 80 mg of Li_5.5PS_4.5Cl_0.8Br_0.7 powder was uniformly dispersed on the opposite side as a separator. The assembly was then subjected to a pressure of 3 tons for 10 s. Following this, a piece of indium (In) foil (3A Materials) with a thickness of 0.1 mm and a diameter of 10 mm was carefully attached to the surface of the Li_5.5PS_4.5Cl_0.8Br_0.7 layer. Then, a 5 mm diameter Li foil (Tianjin Zhongneng Lithium Industry Co., Ltd) was bonded to the In foil at an Li/In weight ratio of 1 : 50. Afterward, the assembled ASSLB was pressed at 1 ton for 10 s, and then placed into a stainless-steel casing (Ningbo Zhengli New Energy Technology Co., Ltd) with a constant applied pressure of 100 MPa. All the preparation processes were conducted in an argon-filled glovebox (H₂O, O₂ < 0.1 ppm). Galvanostatic cycling of the ASSLBs was conducted using a LANBTS-BT-2018R (Hubei, China) cycler at 30 °C.

3 Results

Amorphous Zr-based oxychloride SE for enhanced ion conduction and better contact

The surface morphologies of the cold-pressed pellets synthesized from mechanochemically processed Li₂ZrCl₆ and Li₂ZrCl₄O powers are shown in Fig. 1b–e. The low-magnification scanning electron microscopy (SEM) image in Fig. 1b shows that the Li₂ZrCl₆ pellet is predominantly dense across most regions, which can be attributed to the existence of a substantial amorphous matrix to facilitate particle fusion. However, the co-existing crystalline P3m1 phase (Fig. S1) inside inevitably causes distinct cracks (indicated by arrows in Fig. 1b), accompanied by voids, as observed at a larger magnification (indicated by arrows in Fig. 1c). In comparison, benefitting from the higher amorphous degree, the cold-pressed Li₂ZrCl₄O pellet exhibits a denser surface without obvious cracks or voids (Fig. 1d and e).

Partial oxygen substitution in Li₂ZrCl₆ also brings about enhanced kinetics, as revealed by the ionic conductivity and activation energy results. In Fig. 1f, the Nyquist plots of Li₂ZrCl₆ and Li₂ZrCl₄O SEs at 25 °C indicates the impedance of 156.1 Ω and 51.8 Ω, respectively, based on which the ionic conductivities can be calculated to be 0.51 mS cm⁻¹ for Li₂ZrCl₆ and 1.52 mS cm⁻¹ for Li₂ZrCl₄O (Fig. 1g). Besides higher ionic conductivity, the Arrhenius plots obtained from the temperature-dependent EIS data (Fig. S2a, b and Table S1) for the two SEs from 20 °C to 60 °C in Fig. 1h shows that Li₂ZrCl₄O also possess a lower activation energy of 0.26 eV than that of Li₂ZrCl₆ (Fig. 1i).

As increasing carrier concentration is an effective strategy to elevate the ionic conductivity, where the low-valence In³⁺ is selected to partially substitute Zr⁴⁺ for an enriched Li⁺ content, and thus a series of Li_2+xZr_1−xIn_xCl₄O SEs was synthesized. Fig. 2a depicts the Arrhenius plots obtained from the temperature-dependent EIS data in Fig. S2b–h and Table S1 for Li_2+xZr_1−xIn_xCl₄O with varying stoichiometric ratios of In³⁺ from x = 0.05 to 0.30 tested from 20 °C to 60 °C, and the corresponding ionic conductivities at 25 °C and calculated activation energies are summarized in Fig. 2b. Among these samples, Li_2.15Zr_0.85In_0.15Cl₄O demonstrates the highest ionic conductivity of 2.11 mS cm⁻¹ and lowest activation energy of 0.25 eV, significantly outperforming the other specimens. In addition to the best kinetic performance, Li_2.15Zr_0.85In_0.15Cl₄ also exhibited remarkable mechanical deformability. It is rewarding that In³⁺ substitution does not change the dense morphology of Li_2.15Zr_0.85In_0.15Cl₄O featured by Li₂ZrCl₄O, as shown in Fig. 1d, which shows a similarly well-fused surface on a larger scale (Fig. 2c).

Fig. 2 Structural and electrochemical properties of Li_2+xZr_1−xIn_xCl₄O SEs with varying doping levels. (a) Temperature (T)–dependent ionic conductivities (σ) of the as-prepared Li_2+xZr_1−xIn_xCl₄O SEs. (b) Ionic conductivities at 25 °C and the activation energies (E_a) of Li_2+xZr_1−xIn_xCl₄O SEs. (c) SEM image of the surface of the cold-pressed Li_2.15Zr_0.85In_0.15Cl₄O SE.

An investigation into the effects of ball-milling speed and annealing temperature on ionic conductivity was also conducted. The ionic conductivities of the Li_2.15Zr_0.85In_0.15Cl₄O SE with ball milling speeds of 400 rpm, 600 rpm, and 850 rpm, following by 100 rpm for mixing were calculated to be 1.86 × 10⁻⁴ mS cm⁻¹ for 400 rpm, 8.83 × 10⁻² mS cm⁻¹ for 600 rpm, and 2.11 mS cm⁻¹ for 850 rpm (Fig. S3 and Table S2). In the case of the sample synthesized via ball milling at 850 rpm with the optimal conductivity, annealing at 100 °C, 150 °C, and 200 °C for 2 h led to decreased ionic conductivities of 1.49 mS cm⁻¹, 1.11 mS cm⁻¹, and 0.78 mS cm⁻¹, respectively (Table S3), as calculated from the Nyquist plots in Fig. S4. It seems that high-speed ball milling alone can facilitate a greater degree of amorphization, thus leading to the better ionic conductivity of the Li_2.15Zr_0.85In_0.15Cl₄O SE.

Local structural analysis of Li_2+xZr_1−xIn_xCl₄O

X-ray diffraction (XRD) measurements were carried out to unveil the crucial structural alterations brought about by the aliovalent substitution in Li_2+xZr_1−xIn_xCl₄O (Fig. 3a). No impurities from the raw materials, including Li₂O, ZrCl₄, LiCl, and InCl₃, were found (Fig. S1), indicating the thorough reaction via high-speed ball-milling. In the case of the pristine Li₂ZrCl₄O, the very low peak at approximately 32° can be assigned to P3m1-phase Li₂ZrCl₆ (Fig. S1), which is a common co-existing composition in Zr-based oxychlorides.¹⁴ As the In³⁺ doping amount increases from 0 to 0.05, 0.1 and 0.15, the Li⁺ concentration rises correspondingly, accompanied by the disappearance of low-conductivity Li₂ZrCl₆. This trend correlates with the enhanced ionic conductivity observed at a higher In³⁺ concentration, which is attributed to the increased amorphous degree and carrier concentration. However, when the proportion of In³⁺ exceeds x = 0.20, two minor peaks emerged near 34° and 49°. These peaks were in good agreement with that of Li₃InCl₆ (Fig. S1), which were absent in the samples with x below 0.2, indicating 0.15 as the upper limit of In³⁺ content tolerated by the amorphous Li_2+xZr_1−xIn_xCl₄O matrix. The evolution of the XRD patterns corresponds well with the changing trend of ionic conductivity values, which exhibits a volcano-shaped profile centered at the x = 0.15 phase. The maximum ionic conductivity at x = 0.15 arises from the optimized carrier concentration and preserved amorphous structure, whereas deviations to higher or lower x values degrade the conductivity due to the generation of crystalline Li₃InCl₆ or Li₂ZrCl₆ impurities. Despite the identification of phase evolution upon the Li⁺ content tuning process, the conventional XRD technique is powerless to reveal information on the local structure of amorphous substances, which is important to understand the mechanism of conductivity elevation. Thus, X-ray photoelectron spectroscopy (XPS) analysis and X-ray absorption near-edge structure (XANES) spectroscopy were conducted.

Fig. 3 Local structure characterization of Li_2+xZr_1−xIn_xCl₄O and Li₂ZrCl₆ SEs. (a) XRD patterns of Li_2+xZr_1−xIn_xCl₄O SEs. (b) XPS spectra of O 1s for Li₂ZrCl₆, Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O SEs. (c) XPS spectra of Cl 2p for Li₂ZrCl₆, Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O SEs. (d) XANES of Li₂ZrCl₆ Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O SEs at the Zr K-edge. (e) FT-EXAFS spectra of Li₂ZrCl₆, Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O SEs. (f) Solid-state ⁷Li NMR spectra of Li₂ZrCl_6, Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O SEs.

XPS was conducted on the synthesized Li₂ZrCl₆, Li₃InCl₆, Li₂ZrCl₄O, and Li_2.15Zr_0.85In_0.15Cl₄O, accompanied by purchased Li₂ZrO₃ and In₂O₃ as the reference samples (Fig. S5, S6 and Tables S4–S9). In the XPS spectra of O 1s (Fig. 3b and S6a), three peaks can be fitted and identified. Compared to Li₃InCl₆ without self-contained lattice oxygen (Fig. S6a), the right peak at ∼530.7 eV in the O 1s spectra of Li₂ZrO₃ and In₂O₃ with abundant self-contained lattice oxygen presents an obviously stronger intensity, and thus assigned to the characteristic metal–oxygen (In/Zr–O) bonds. This phenomenon also occurs in the series of XPS patterns of Li₂ZrCl₆, Li₂ZrCl₄O, and Li_2.15Zr_0.85In_0.15Cl₄O, where the O 1s peaks assigned to Zr–O in Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O are stronger than that of Li₂ZrCl₆ (Fig. 3b). Considering that no Li₂O or other common Zr/In-based oxides (e.g., ZrO₂ and In₂O₃) was detected in the series of XRD patterns of Li_2+xZr_1−xIn_xCl₄O in Fig. 3a, the difference in O 1s spectra in Fig. 3b indicates the successful formation of metal–oxygen bonds in the amorphous matrix of Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O by replacing LiCl with Li₂O as the lithium source. It should be noted that the metal–oxygen O 1s peaks in the Li₂ZrCl₆ and Li₃InCl₆ samples without oxygen in their standard formulas arise from the inevitable moisture absorption during storage and transfer. Furthermore, Li₂ZrCl₆ exhibits a stronger metal–oxygen O 1s peak than Li₃InCl₆ due to the stronger hydrolysis trend of Zr-based chloride to irreversibly react with moisture (ZrCl₄ + H₂O → ZrO₂ + HCl↑), while In-based chloride mainly combines with water to become hydrated (InCl₃ + H₂O → InCl₃·xH₂O), which is reversible and milder.

The left peaks at 533.44–534.58 eV and middle peaks at 532.09–532.44 eV, shared by all the samples, are attributed to the surface-adsorbed O–H or C–O impurities. Meanwhile, the Cl 2p XPS spectra, with similar positions (2p_3/2 peak at 198.68–198.98 eV) and shape across all the Cl-contained samples, including Li₂ZrCl₆, Li₂ZrCl₄O, Li_2.15Zr_0.85In_0.15Cl₄O and Li₃InCl₆, confirm the existence of metal-chloride bonds in the Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O samples (Fig. 3c, S6d and Tables S4–S7), but no more information be provided by the very similar Zr 3d and In 3d XPS spectra among the Zr-containing and In-containing samples except for the oxidation state of Zr⁴⁺ and In³⁺ (Fig. S6b and c). Considering only the limited local structure information revealed by XPS, XANES spectroscopy was utilized to further investigate the more-detailed local coordination environments in the Zr-based amorphous matrix (Fig. 3d and e).

Specifically, Zr K-edge XANES measurements were performed to confirm the average local chemistry. This choice is grounded in the fact that the line shape of the K-edge of a transition metal is significantly influenced by the nearest interatomic distances of adjacent atoms. As depicted in Fig. 3d, the Li₂ZrCl₆ SE exhibited distinct splits. These can be construed as manifestations of fragmented hexagonal close-packed (hcp)-Li₂ZrCl₆ structures with varying bond lengths, typically differing by more than 0.18 Å. In stark contrast, the O-substituted Li₂ZrCl₄O sample presented a smoother profile, indicating that the differences in bond lengths within the Zr–Cl/O polyhedra had diminished. Upon the incorporation of In³⁺, the Zr K-edge spectra of the Li_2.15Zr_0.85In_0.15Cl₄O SE became even smoother, with no discernible splitting. This further implies that in the amorphous Li_2.15Zr_0.85In_0.15Cl₄O SE, the nearest Zr-related bond lengths were more regular and homogeneous, suggesting that the incorporation of In³⁺ and O²⁻ induced localized structural changes. In addition, Zr K-edge Fourier transform-extended X-ray absorption fine structure (FT-EXAFS) spectroscopy was employed to quantitatively analyse the coordination environment surrounding the central Zr atom in the Li₂ZrCl₆, Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O SEs (Fig. 3e and S7a). By leveraging single-scattering data from ZrCl₄ (at 2.1 Å) and ZrO₂ (1.5 Å), the peaks generated by the O and Cl scatters around Zr were accurately identified. The results showed an increase in the intensity associated with the Zr–O interactions and a decrease in that of Zr–Cl interactions. This phenomenon clearly demonstrates that O was successfully introduced into the Li_2.15Zr_0.85In_0.15Cl₄O SE.

Furthermore, the wavelet-transformed (WT) EXAFS spectra of the Li₂ZrCl₆, Li₂ZrCl₄O and Li_2.15Zr_0.85In_0.15Cl₄O SEs, as presented in Fig. S7b–d, respectively, provide compelling evidence for the coordination environments of Zr–O and Zr–Cl. According to the quantitative fitting results of Zr K-edge EXAFS for the three samples (Table S10), different from the confirmed [ZrCl₆] local structure in Li₂ZrCl₆, the coordinating conditions of Zr in the Zr-based oxychlorides are O/Cl = 1.9/3.9 for Li_2.15Zr_0.85In_0.15Cl₄O and O/Cl = 1.7/4.0 for Li₂ZrCl₄O. Despite the uncertainty to completely identify the local structures in the amorphous matrix, taking into consideration the O/Cl ratio of approximately 2/4, which is similar to that of LiNb(Ta)OCl₄,¹⁹ we tend to speculate the six-coordinated octahedron [Zr/InO₂Cl₄] as the dominant structure units bridged via oxygen, accompanied by various Zr/In-centered oxychloride polyhedrons [Zr/InO_aCl_b].¹⁴ Besides the above-mentioned bridging oxygen in the amorphous matrix, oxygen also exists in the non-bridging forms, which plays a crucial role in creating a relatively open framework for faster Li⁺ conduction.²⁰

The unique local structures with highly amorphous degree accelerate Li⁺ transport, as proven by the obviously narrowed peaks in the solid-state ⁷Li NMR spectra of Li_2.15Zr_0.85In_0.15Cl₄O and Li₂ZrCl₄O compared to that of Li₂ZrCl₆ (Fig. 3f).²⁰

ASSLBs based on Li_2.15Zr_0.85In_0.15Cl₄O SEs

The linear sweep voltammetry (LSV) measurement conducted from 2.0 to 5.0 V vs. Li⁺/Li indicates the oxidative stability of Li_2.15Zr_0.85In_0.15Cl₄O up to 4.28 V vs. Li⁺/Li, which is superior to that of 4.18 V vs. Li⁺/Li for Li₂ZrCl₆ and 4.23 V vs. Li⁺/Li for Li₂ZrCl₄O (Fig. S8), thus facilitating the stable cycling of 4 V-class cathodes. The electrochemical performances of the composite cathodes employing Zr-based (oxy)chloride SEs and NCM92 were systematically investigated in ASSLBs, in which the higher ionic conductivity of the Li_2.15Zr_0.85In_0.15Cl₄O SE than that of Li₂ZrCl₆ is expected to exhibit lower polarization and better capacity delivery especially under high rates.

The ASSLB employing the Li₂ZrCl₆ SE demonstrates a poor rate capability with low discharge capacities of 202.6, 201.0, 188.9, 169.0, 127.5, 44.9, and 0.4 mAh g⁻¹ at rates of 0.2, 0.4, 0.8, 2.0, 4.0, 8.0, and 20C, respectively (Fig. 4a). In contrast, Li_2.15Zr_0.85In_0.15Cl₄O SE enables an obviously improved rate performance with 214.9 mAh g⁻¹ at 0.2C, 209.9 mAh g⁻¹ at 0.4C, 200.4 mAh g⁻¹ at 0.8C, 191.1 mAh g⁻¹ at 2.0C, 175.3 mAh g⁻¹ at 4.0C, 125.6 mAh g⁻¹ at 8.0C, and 69.3 mAh g⁻¹ at 20C (Fig. 4b). The capacities of the cathodes using Li₂ZrCl₆ and Li_2.15Zr_0.85In_0.15Cl₄O exhibit obviously enlarged differences when the rate reaches 2.0C or higher (Fig. 4c), the trend of which can be viewed in the capacity retention summary based on 0.2C at different rates (Fig. 4d). Taking into consideration the remarkable capacity retention of 81.57% at 4.0C compared to that at 0.2C for the Li_2.15Zr_0.85In_0.15Cl₄O-based ASSLB, a long-cycling stability test was then conducted at the same rate. Aside from higher capacity delivery, Li_2.15Zr_0.85In_0.15Cl₄O also enables the ASSLB to stably cycle for 450 cycles at 4.0C with 91.44% capacity retention, which is superior to that of 88.61% by the Li₂ZrCl₆-based ASSLB (Fig. 4e).

Fig. 4 Electrochemical performance of Li₂ZrCl₆ and Li_2.15Zr_0.85In_0.15Cl₄O SEs at 25 °C under 4.3 V. (a) Charge–discharge curves at different rates for ASSLBs using Li₂ZrCl₆ SE. (b) Charge–discharge curves at different rates for ASSLBs using Li_2.15Zr_0.85In_0.15Cl₄O SE. (c) Rate performances of Li₂ZrCl₆ and Li_2.15Zr_0.85In_0.15Cl₄O-based ASSLBs from 0.2 to 20C. (d) Capacity retention of Li₂ZrCl₆ and Li_2.15Zr_0.85In_0.15Cl₄O-based ASSLBs at different C-rates based on 0.2C. (e) Long-term cycling of Li₂ZrCl₆ and Li_2.15Zr_0.85In_0.15Cl₄O-based ASSLBs at 4.0C.

An increased charge upper limit of 4.8 V vs. Li⁺/Li enables higher discharge capacities of 220.6 mAh g⁻¹ and 240.5 mAh g⁻¹ at 0.2C for the ASSLBs based on Li₂ZrCl₆ and Li_2.15Zr_0.85In_0.15Cl₄O (Fig. 5a and b), respectively. However, with a higher rate of 4.0C, the Li_2.15Zr_0.85In_0.15Cl₄O-based ASSLB exhibits a less increased polarization and a higher capacity of 203.0 mAh g⁻¹ at 4.0C than that of Li₂ZrCl₆ (169.8 mAh g⁻¹) (Fig. 5a and b). In the following long-cycling tests at 4.0C, after 250 cycles, Li_2.15Zr_0.85In_0.15Cl₄O also enables a higher capacity retention of 81.86% compared to the Li₂ZrCl₆-based ASSLB (81.44%) (Fig. 5c).

Fig. 5 Electrochemical performances and morphologies of composite cathodes based on Li₂ZrCl₆ and Li_2.15Zr_0.85In_0.15Cl₄O SEs. (a) Charge–discharge curves at different rates for Li₂ZrCl₆ SE. (b) Charge–discharge curves at different rates for Li_2.15Zr_0.85In_0.15Cl₄O SE. (c) Long-term cycling of Li₂ZrCl₆ and Li_2.15Zr_0.85In_0.15Cl₄O-based ASSLBs at 4.0C. (d) FIB-SEM images of the composite cathodes before and after 50 cycles.

Although Li_2.15Zr_0.85In_0.15Cl₄O exhibits lower polarization and higher capacity delivery under the same rates due to its better ion conduction than Li₂ZrCl₆, it should be noted that the differences in the aspect of long-cycling capacity retention do not seem to be as obvious as the polarization and capacity delivery at high rates (Fig. 5a and b). On the one hand, less polarization and higher capacity delivery also lead to a deeper charge/discharge status and practical charge potential closer to the set upper limit value, which may cause faster degradation for both the cathode particles and electrolyte/cathode interfaces. On the other hand, Li₂ZrCl₆ itself is low crystallized due to the intense ball milling,¹² and thus easily deformable in the composite cathode. Nevertheless, Li_2.15Zr_0.85In_0.15Cl₄O with a higher amorphous degree enables better deformability and more intimate solid–solid contact in the cathode before and during cycling at 4.3 or 4.8 V (Fig. 5d), improving the condition of voids generated at the electrolyte/cathode interface (yellow arrows in Fig. 5d), especially under the high upper charge limit of 4.8 V, as revealed by the focused ion beam-scanning electron microscope (FIB-SEM) images.

4 Conclusions

In conclusion, we successfully synthesized an Li_2.15Zr_0.85In_0.15Cl₄O SE through a one-step ball-milling process. This material features a dual-anion oxychloride sublattice and a highly amorphous matrix. It demonstrates remarkable ionic conductivity, reaching up to 2.11 mS cm⁻¹, along with notable cost-effectiveness. By integrating the analytical data from XANES analysis, it was discovered that the incorporation of O²⁻ augmented the amorphous content and enhanced the ionic conductivity within the oxychloride framework. The amorphous Li_2.15Zr_0.85In_0.15Cl₄O SE also demonstrated excellent compatibility with high-nickel single-crystal cathode materials (NCM92), delivering outstanding electrochemical performances at both 4.3 V and 4.8 V vs. Li⁺/Li.

Author contributions

Shiliang Yang: formal analysis, data curation, and investigation. Qilin Feng: project administration, validation, and writing-original draft. Xiaoke Liu: formal analysis and data curation. Changmeng Xu: investigation. Xianyi Liu: formal analysis. Wenxin He: formal analysis. Jiangmin Jiang: writing-review and editing. Tao Ma: resources, formal analysis, funding acquisition, writing-review and editing. Jipeng Fu: resources, formal analysis, funding acquisition, writing-review and editing. Yichen Yin: conceptualization, supervision, methodology, funding acquisition, investigation, writing-review and editing.

Conflicts of interest

There are no conflicts of interest to declare.

Data availability

All data needed to support the conclusions in the paper are presented in the manuscript and the SI.

Supplementary information: Additional figures and tables (PDF). See DOI: https://doi.org/10.1039/d5ta05340d.

Acknowledgements

We acknowledge the financial support from the National Natural Science Foundation of China (No. 22475235 and 62205322), the Startup Research Fund of China University of Mining and Technology (102525017), the Startup Research Fund of Chaohu University (Grants KYQD-2023005), the Key Projects in the 2025 Annual Plan of China Jiliang University (HEX2025019), the National College Students' Innovation and Entrepreneurship Training Program (Grant No. 202410356069) and the Anhui Province University Natural Science Research Project (Grants 2023AH030093). We thank the Anhui Absorption Spectroscopy Analysis Instrument Co, Ltd for XAFS measurements and analysis.

References

1. X. G. Han, Y. H. Gong, K. Fu, X. F. He, G. T. Hitz, J. Q. Dai, A. Pearse, B. Y. Liu, H. Wang, G. Rublo, Y. F. Mo, V. Thangadurai, E. D. Wachsman and L. B. Hu, Nat. Mater., 2017, 16, 572–579.
2. W. Luo, Y. H. Gong, Y. Z. Zhu, Y. J. Li, Y. G. Yao, Y. Zhang, K. Fu, G. Pastel, C. F. Lin, Y. F. Mo, E. D. Wachsman and L. B. Hu, Adv. Mater., 2017, 29, 1606042.
3. R. Murugan, V. Thangadurai and W. Weppner, Angew. Chem., Int. Ed., 2007, 46, 7778–7781.
4. W. Luo, Y. Gong, Y. Zhu, K. K. Fu, J. Dai, S. D. Lacey, C. Wang, B. Liu, X. Han and Y. Mo, J. Am. Chem. Soc., 2016, 138, 12258–12262.
5. N. Kamaya, K. Homma, Y. Yamakawa, M. Hirayama, R. Kanno, M. Yonemura, T. Kamiyama, Y. Kato, S. Hama, K. Kawamoto and A. Mitsui, Nat. Mater., 2011, 10, 682–686.
6. Y. Kato, S. Hori, T. Saito, K. Suzuki, M. Hirayama, A. Mitsui, M. Yonemura, H. Iba and R. Kanno, Nat. Energy, 2016, 1, 16030.
7. Y. X. Li, S. B. Song, H. Kim, K. Nomoto, H. Kim, X. Y. Sun, S. Hori, K. Suzuki, N. Matsui, M. Hirayama, T. Mizoguchi, T. Saito, T. Kamiyama and R. Kanno, Science, 2023, 381, 50–53.
8. S. Wang, Y. J. Wu, T. H. Ma, L. Q. Chen, H. Li and F. Wu, ACS Nano, 2022, 16, 16158–16176.
9. Y. Ishiguro, K. Ueno, S. Nishimura, G. Iida and Y. Igarashib, Chem. Lett., 2023, 52, 237–241.
10. L. D. Zhou, T. T. Zuo, C. Y. Kwok, S. Y. Kim, A. Assoud, Q. Zhang, J. Janek and L. F. Nazar, Nat. Energy, 2022, 7, 83–93.
11. 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.
12. K. Wang, Q. Y. Ren, Z. Q. Gu, C. M. Duan, J. Z. Wang, F. Zhu, Y. Y. Fu, J. P. Hao, J. F. Zhu, L. H. He, C. W. Wang, Y. Y. Lu, J. Ma and C. Ma, Nat. Commun., 2021, 12, 4410.
13. J. Z. Wang, F. Chen, L. Hu and C. Ma, Nano Lett., 2023, 23, 6081–6087.
14. S. M. Zhang, F. P. Zhao, L. Y. Chang, Y. C. Chuang, Z. Zhang, Y. M. Zhu, X. G. Hao, J. M. Fu, J. T. Chen, J. Luo, M. S. Li, Y. J. Gao, Y. N. Huang, T. K. Sham, M. D. Gu, Y. P. Zhang, G. King and X. L. Sun, J. Am. Chem. Soc., 2024, 146, 2977–2985.
15. B. Ravel and M. Newville, J. Synchrotron Radiat., 2005, 12, 537–541.
16. M. Newville, J. Synchrotron Radiat., 2001, 8, 322–324.
17. C. J. Lou, W. D. Zhang, J. Liu, Y. A. Gao, X. Sun, J. P. Fu, Y. C. Shi, L. G. Xu, H. J. Luo, Y. J. Chen, X. Gao, X. J. Kuang, L. Su and M. X. Tang, Chem. Sci., 2024, 15, 3988–3995.
18. J. C. Liu, J. Liu, M. X. Tang, J. P. Fu, X. J. Kuang and J. W. Ma, Adv. Funct. Mater., 2024, 34, 11.
19. Y. Tanaka, K. Ueno, K. Mizuno, K. Takeuchi, T. Asano and A. Sakai, Angew. Chem., Int. Ed., 2023, 62, e202217581.
20. Z. Li, Y. Mu, K. Lu, G. Kang, T. Yang, S. Huang, M. Wei, L. Zeng and Y. Li, Angew. Chem., Int. Ed., 2025, 64, e202501749.

---

Footnote

† Shiliang Yang, Qilin Feng and Xiaoke Liu contributed equally to this work.

---