Introduction

The need for efficient and affordable energy storage devices to manage intermittent renewable energy has kindled immense interest in functional perovskite materials, representing a distinctive class of ionic crystals comprising several oxide and halide structures with similar octahedral features1,2,3,4. ABO3 (A and B are divalent and tetravalent metallic cations, respectively) structured oxide perovskites are known as promising ferroelectric, dielectric, magnetic, and energy storage materials5,6,7. Owing to the abundance of oxygen vacancies, oxide perovskites have been successfully applied as catalytic electrode materials8,9,10,11. Instead, halide perovskites (ABX3, A is monovalent cation, B is divalent cation, and X is I, Br, or Cl) possess relatively narrower bandgaps due to the large difference in electronegativity between halogens and oxygen, thereby used for optoelectronic applications such as the light absorber and emissive layer in solar cells and light-emitting diodes, respectively12,13,14,15. Nevertheless, the study of halide perovskites in batteries has reached a rudimentary stage, though being contained to a subsidiary role in storage applications.

The research on molecular-level low-dimensional (LD) crystals such as two-dimensional (2D), one-dimensional (1D), and zero-dimensional (0D) perovskites are advanced in the meanwhile due to demand for structural stability16,17,18,19, some of which reflect deficient formation energy at ambient conditions being suitable for multifunctional purposes20,21,22. Among them, the LD organic-inorganic hybrid perovskites demonstrate a substantial quantum and dielectric confinement due to the breakdown of three-dimensional (3D) frameworks as the inclusion of insulating organic moieties23. This special lattice ordering promotes improved structural stability and expanded bandgaps that bring out an exciting intersection of the halide and oxide perovskites’ applications, including ferroelectricity and energy storage devices24,25,26,27,28,29,30,31,32.

Unlike ancillary oxide perovskites that provide oxygen vacancies for metal-air batteries33,34, halide perovskites perform as the halogen reservoir and restrain the shuttle effect by using Van Der Waals forces (hydrogen bonds and halogen bonds) and steric hindrance from A-site organics, eventually realizing a three-electron transfer process35. However, those encouraging attempts to develop halide perovskite cathodes failed to avail the full range of perovskite materials given that the redox reaction of the B-site cations sits electrochemical inert during the conversion process (of halogens). This undermines the whole discharge capacity. In this regard, replacing those noble B-site metal cations with tetravalent chalcogenide cations to construct chalcogen-halide octahedra is expected to offer full utilization of cathode materials and guarantee a reliable multiple electron transfer. Nevertheless, the high-valent hexavalent and tetravalent chalcogen cations are neither stable nor electrochemically active in aqueous electrolytes, causing irreversible redox processes in batteries36,37,38,39,40.

Herein, we designed a benzyltriethylammonium tellurium iodide perovskite, (BzTEA)2TeI6, as the cathode material for problem-solving and demonstrated its special chemical processes in aqueous zinc ion batteries. In our design, the high charge density surrounding each tetravalent Te4+ cation causes the formation of the tellurium-iodide octahedron unit, which is embedded in the A-site organic ligands (BzTEA) matrix and supplies high elemental iodine and tellurium content of over 71 wt.%. The robust Van Der Waals forces such as Te-I…Cl-Te and Te-I…I halogen bonds on the perovskite surface promote the localization of active elements and avert the undesired shuttling of oxidative polyiodide and tellurium polychloride ions, as supported by the density functional theory (DFT) calculations. Consequently, the Zn||(BzTEA)2TeI6 battery delivers a highly reversible eleven-electron transfer mode on account of the two-electron I+/I0/I- and eight-electron Te6+/Te4+/Te0/Te2- redox reactions, and one-electron Cl0/Cl- transfer gained from the chloride electrolyte, all of which favor a high discharge capacity up to 473 mAh g-1Te/I and a large energy density of 577 Wh kg-1Te/I at 0.5 A g-1 represented by five prominent voltage plateaus at 1.81 V, 1.64 V, 1.53 V, 1.26 V, and 0.51 V. The feasibility of the (BzTEA)2TeI6 perovskite cathode is further verified by the zinc pouch cells.

Results

(BzTEA)2TeI6 perovskite cathode

Chalcogens, such as sulfur, selenium, and tellurium, display a broad spectrum of valence states (−2, 0, +2, +4, +6) and exhibit significant redox potentials (Fig. 1a). This characteristic theoretically renders them suitable for energy storage, while the stability of high-valent chalcogen cations remains a crucial consideration. Halogens, on the other hand, are proven effective in giving high redox potentials in aqueous zinc ion batteries41,42. A proper integration of chalcogen and halogen chemistry should promote problem-solving and actualize high-energy zinc ion batteries. Implementing halogen redox in halide perovskites brings our attention to potential structural design and prompts the exploration of chalcogen halide perovskites as cathode materials. Actually, the typical ABX3 perovskites consist of corner-sharing [BX6]4- octahedra and offset of A-site atoms in octahedrons cavities throughout the whole 3D matrix43. In contrast, the molecular-level LD perovskite materials crystalize in a way where the [BX6]4- octahedron unit is separated by A-site cations in specific directions and upholds a high structural tunability (Supplementary Fig. 1)44,45. As such, we have now aim to replace conventional electrochemical inert B-site cations with tetravalent chalcogenide cations while retaining the octahedral BX6 motif to enable the formation of A2BIVX6 vacancy-ordered perovskites46. This change is expected to maximize the utilization of perovskite materials by enabling multivalent reactions (B2-/B0/B4+/B6+) of both chalcogen and halogen elements, as depicted in Fig. 1a. The special lattice arrangement ensures the confinement of chalcogen and halogen elements in the same perovskite structure, which creates the platform for chalcogen- and halogen-related redox reactions for high-energy batteries. As summarized in Supplementary Table 1, the cogitation theoretically upholds high redox potential and multiple electron transfer associated with B-site chalcogen and X-site halogens of the perovskite cathode materials, attempting to overcome the deficiency of conventional chalcogen cathodes and enable the electrochemically inert high-valent chalcogen redox47,48. As a proof-of-concept, (BzTEA)2TeI6 perovskite, which possesses a high proportion of active elements (>71 wt.%), is thus proposed as conversion-type cathodes for aqueous zinc ion batteries with a special eleven-electron transfer process. (Fig. 1b).

Fig. 1: Chalcogen halide perovskite structure design and electrochemical properties.
figure 1

a Redox potential of chalcogens and halogens. b Structural arrangement of low-dimensional (BzTEA)2TeI6 perovskite where tellurium and iodine elements sit as the B- and X-site components, respectively. The CV curves of the perovskite cathode in c 2 M ZnSO4 (first two cycles), d 30 M ZnCl2, and e Ch0.4Zn0.6Cl1.6·1.5H2O.

Full size image

Regardless of the lattice arrangement at molecular levels, (BzTEA)2TeI6 microcrystals prepared by a modified saturation recrystallization method crystallized into bulk rod shape with an average length of less than 50 µm, as illustrated in the scanning electron microscope (SEM) image in Supplementary Fig. 219. The perovskite microcrystals delivered identical XRD diffractions with the theoretical simulation in Supplementary Fig. 3 and appeared black due to its narrow optical bandgap, aligning with the ultraviolet absorption spectra in Supplementary Fig. 4. SEM-mapping in Supplementary Fig. 5 further revealed the even elemental distribution of (BzTEA)2TeI6 microcrystals associated with 14.5 at.% of Te and 85.5 at.% of I according to the energy-dispersive X-ray spectroscopy (EDS) test. Fourier Transform Infrared (FTIR) spectra in Supplementary Fig. 6 indicated the protonation of A-site N-H bond in (BzTEA)2TeI6. Thermogravimetric analysis (TGA) and derivative thermogravimetry (DTG) curves in Supplementary Fig. 7 confirmed the structural stability and hydrophobic feature of the tellurium iodide perovskite according to the negligible weight loss below 211 °C. Raman spectra in Supplementary Fig. 8 reveal symmetric A1g and asymmetric Eg stretching of Te-I bond at approximately 153 cm-1 and 107 cm-1 in (BzTEA)2TeI6, resembling the Te-O bond observed in TeO249.

Tellurium and halogen redox

The halide perovskite cathode was first examined in Zn||2 M ZnSO4||(BzTEA)2TeI6 Swagelok cells and achieved with two redox peaks, corresponding to I0/I- and Te4+/Te0, the latter of which experienced fast decay of current according to the cyclic voltammetry (CV) curves in Fig. 1c. In comparison, 2 M Zn(OTF)2, 2 M Zn(OAc)2, and 15 M ZnCl2 electrolytes all failed to activate the redox reaction of high-valent tellurium cations (Supplementary Figs. 9, 10). 30 M ZnCl2, containing reduced amount of free water, was then used as the electrolyte with an attempt to stabilize the tetravalent tellurium cations50. Surprisingly, it enabled the perovskite cathode with an additional Te6+/Te4+ redox at around 1.5 V. It should be noted that the Te6+/Te4+ redox was utterly absent in the “TeO2 + ZnI2” and “Te + I2” cathodes (Fig. 1d and Supplementary Fig. 11), which reveals the special structure of (BzTEA)2TeI6 for activation of the new redox. The observation also suggested that the activity of water and chloride anions may play an essential role in stabilizing high-valent tellurium cations. Based on that, we introduced choline chloride (ChCl) into the concentrated ZnCl2 electrolyte to devise the Ch0.4Zn0.6Cl1.6 · 1.5H2O electrolyte, which was found to have limited water activity and enough mobility of chloride ions to coordinate with high-valent tellurium cations during cycling. As shown in Fig. 1e, the activity endowed Zn||Ch0.4Zn0.6Cl1.6 · 1.5H2O||(BzTEA)2TeI6 prominent discharge peaks, including Te6+/Te4+ and Te4+/Te0 redox pairs at 1 mV. In sharp contrast, both “TeO2” and “Te + I2” cathodes failed to enforce the Te6+/Te4+ redox even when coupled with the same modified electrolyte, again highlighting the special chemical design of (BzTEA)2TeI6 (Supplementary Fig. 12).

In order to figure out the critical role of (BzTEA)2TeI6 perovskite structure for the rich chemistry, we subsequently conducted detailed electrochemical studies of these batteries. As shown in Fig. 2a, the “TeO2 + ZnI2” cathode presented two types of redox, including I0/I+/I- and Cl0/Cl-, which experienced an unusual attenuation as the scan rate increased, together with a mild Te6+/Te4+ redox reaction probably due to the catalytic role of iodine atom (Supplementary Fig. 13). The (BzTEA)2TeI6 cathode, on the contrary, featured four sharp discharge peaks from the Cl0/Cl-, I+/I0, Te6+/Te4+, and Te0/Te2- redox pairs and one broad peak due to the overlay of Te4+/Te0 and I0/I- as will be discussed later (Fig. 2b and Supplementary Fig. 14). Further, as illustrated in Fig. 2c, the surface-controlled process of the battery based on perovskite cathodes gradually grew from 21.5% at 0.5 mV s-1 to 40.2% at 3 mV s-1. In comparison, that for the “TeO2 + ZnI2” cathode started from 11.1% at 0.5 mV s-1 to 23.5% at 3 mV s-1 due to weak adsorption of active elements at a high scan rate (Supplementary Fig. 15). The derived b value of each cathodic peak generally fell between 0.5 and 1.0 and exhibited a combined action of the faradic and capacitive processes during the conversion reactions (Fig. 2d)51. However, the b values of the control sample were far below 0.5 and turned negative for Cl0/Cl-, suggesting the poor cycling stability of the “TeO2 + ZnI2” cathode and uncontrolled loss of chlorine elements (Supplementary Figs. 15, 16). The summary in Fig. 2e demonstrated their difference in b values. It highlighted the importance of (BzTEA)2TeI6 and ChCl for properly operating batteries, aside from taking full advantage of the B-site tellurium elements. As a comparison, the Zn||(BzTEA)2TeI6 battery in 30 M ZnCl2 experienced severe degradation under identical test conditions, even though it delivered a Te0/Te2- redox pair with a larger b value (0.6) and a mild Te6+/Te4+ redox pair (Supplementary Fig. 17). Figure 2f presents a typical galvanostatic discharge curve of Zn||(BzTEA)2TeI6 battery at 0.5 A g-1, which brought out an eleven-electron transfer process, including three-electron transfer from the iodine element, eight-electron transfer from tellurium element and one from chlorine element, following the result from the differential capacity (dQ/dV) plot in Supplementary Fig. 18.

Fig. 2: Electrochemical properties of the Zn||(BzTEA)2TeI6 battery.
figure 2

CV curves of batteries coupled with a “TeO2 + ZnI2” cathode and b “(BzTEA)2TeI6” cathode in Ch0.4Zn0.6Cl1.6·1.5H2O electrolyte. c Diffusion-/surface-controlled contribution at different scan rates and d the fitting plots between log(i) and log(v) of the cathodic peaks, and e the derived b value a summary diagram. f Discharge curve at a current density of 0.5 A g−1 demonstrates the redox reactions of tellurium and halogen elements.

Full size image

Mechanism of the eleven-electron transfer

The Raman spectra recorded during the charge and discharge process indicated the presence of I5- at 169 cm-1, consistent with theoretical calculations, while signals from high-valence tellurium cations were indistinct (Supplementary Fig. 19). This may be due to the overlap of Te6+ signal with the [ZnCl2+x(H2O)y] cluster at around 300 cm-1 and the Te4+ signal below 100 cm-1 being affected by intense Rayleigh scattered laser light (noise)33,40,52,53. The XRD and its enlarged patterns in Supplementary Fig. 20 reveal a broad peak around 10°, close to the main diffraction peaks of (BzTEA)2TeI6 perovskites, whereas the blank electrode consisting solely of Ketjen black and PVDF binder exhibits no characteristic peak at that position. X-ray photoelectron spectroscopy (XPS) measurements were further performed to evaluate the perovskite cathode’s conversion reactions. As demonstrated in Fig. 3a, Te0 and Te2- dominated at discharge states; the signal of Te4+ first appeared at 576 eV when charged to 1.2 V and shifted to 576.9 eV, verifying the formation of Te6+. The presence of Te-O at 579.2 eV is attributed to the oxidation of high-valent tellurium ions in damp air. The Cl 2p3/2 core level spectra in Fig. 3b exhibit the signals from lattice Cl- changing from 198.9 eV at 0.5 eV, 199.4 eV at 1.7 V, to 199.6 eV at 1.9 V, contributing to an energy span of over 0.7 eV. Instead, the adsorbed Cl- changed from 198.2 eV at 0.5 V to 198.7 eV at 1.7 V and slightly increased to 198.73 eV at 1.9 V, corresponding to an energy change of 0.5 eV. The energy evolution of the two types of Cl- revealed that a certain degree of (dynamic) halide exchange between iodide and chloride ions could promote a reliable Cl0/Cl- redox reaction, consistent with the CV performance discussed above. The I 3d3/2 core level spectra in Fig. 3c demonstrate the conversion from I+ at 1.7 V, to I0 at 1.3 V and I- at 0.5 V, verifying the successful operation of the I+/I0/I- redox pairs54.

Fig. 3: Characterization of cycled (BzTEA)2TeI6 perovskite cathode.
figure 3

XPS spectra of a Te 3d, b Cl 2p, and c I 3d3/2 core levels.

Full size image

Molecular dynamics (MD) simulations and DFT calculations were subsequently conducted to understand the working principle of the chalcogen halide perovskite cathode. As depicted in Fig. 4a, the ZnCl2 · 1.4H2O electrolyte was dominated by tetrahedral [ZnCl4]2- and octahedral [ZnCl4(H2O)2]2- clusters. Notably, as the inclusion of ChCl, the coordinated water molecules surrounding Zn2+ increased in the form of [ZnCl3(H2O)3]- and octahedral [ZnCl2(H2O)4] cluster, giving rise to the free Cl- radicals that sat in the electrolyte voids (Supplementary Fig. 21). The mean squared displacement (MSD) study in Fig. 4b proved a much higher mobility of Cl- ion in Ch0.4Zn0.6Cl1.6 · 1.5H2O (4.30 ×10-8 cm2 s-1) while that for ZnCl2 · 1.4H2O was 8.64 ×10-9 cm2 s-1. The radial distribution function (RDF) result in Fig. 4c confirmed that the coordination number of Cl- decreased from 3.6 to 3.3 after introducing ChCl. Moreover, the amorphous features of Ch0.4Zn0.6Cl1.6 · 1.5H2O and the absence of undissolved solutes emphasized that the designed electrolyte was a homogeneous solution (Supplementary Figs. 22, 23), which offered high Cl- mobility to compensate and stabilize high-valent tellurium ions.

Fig. 4: Theoretical studies of the eleven-electron transfer.
figure 4

a Visualization of ZnCl2·1.4H2O (left) and Ch0.4Zn0.6Cl1.6·1.5H2O (right) electrolytes. b MSD and c RDF spectra derived from MD simulations. d Migration energy of chloride ions on TeO2 and perovskite surface. e Formation energy of associated redox pairs in vacuum and on perovskite surface. f Surface adsorption energy of (BzTEA)2TeI6 toward intermediate tellurium and halogen elements. Inset in f denotes the coordination between perovskite and TeCl5+ ions.

Full size image

Of note, the (BzTEA)2TeI6 perovskite surface synergistically provided a fast channel for transporting Cl- ions, favoring the stabilization of high-valent tellurium ions. As shown in Fig. 4d, the migration energy of Cl- along the (1 0 −1) lattice plane of perovskite oscillated from −9.5 to 9.6 eV. In contrast, the amplitude of the migration barrier on TeO2 soared to over 156 eV. In addition to kinetic influences, the conversion reactions on the (BzTEA)2TeI6 surface were thermodynamically preferred. Specifically, as demonstrated in Fig. 4e, the formation energy of critical chalcogen- and halogen-related redox reactions all decreased, especially for TeCl3+ and TeCl5+, which reduced from 7.7 eV to 5.8 eV and 3.0 eV to 0.8 eV, respectively. Besides, the presence of perovskite surface ended up enhancing the confinement ability toward chalcogen and halogen elements by virtue of halogen bonds (such as Te-I…Cl-Te and Te-I…I) and positive dangling bonds, as summarized in Fig. 4f. The effective coordination among perovskite surface, high-valent cations, and passivating Cl- ligand jointly favored their stabilization and associated redox reactions (Supplementary Figs. 24, 25). The more significant adsorption energy of I5- (−1.97 eV) than I3- (−1.30 eV) helped explain the proliferation of I5- during the charging process, pointing out improved reaction kinetics and suppressed shuttle effects (Supplementary Fig. 26).

Electrochemical performance of Zn||(BzTEA)2TeI6 battery

We subsequently assembled batteries to examine the practical performance of the perovskite cathode in the Ch0.4Zn0.6Cl1.6 · 1.5H2O electrolyte. As presented in Fig. 5a, the Zn||(BzTEA)2TeI6 battery delivered a steady capacity of 473 mAh g-1Te/I at 0.5 A g-1 and 258 mAh g-1Te/I at 3 A g-1, which reverted to 411 mAh g-1Te/I after current reset, showing a high capacity retention of 87%. The relatively lower coulombic efficiency (CE) at lower C-rates could be attributed to the loss of chlorine gas during the slow charge/discharge process (Supplementary Fig. 27). This loss was suppressed at higher C-rates due to the shorter cycle time and saturated Cl₂ dissolution in the electrolytes. The corresponding galvanostatic charge-discharge (GCD) curves in Fig. 5b show five stable discharge plateaus at 1.81 V (for Cl0/Cl-), 1.64 V (for I+/I0), 1.53 V (for Te6+/Te4+), 1.26 V (for overlayed Te4+/Te0 and I0/I-), and 0.51 V (for Te0/Te2-), outperforming the Zn||Ch0.4Zn0.6Cl1.6·1.5H2O||“TeO2 + ZnI2” and Zn||30 M ZnCl2||(BzTEA)2TeI6 batteries (Supplementary Fig. 28) The highly reversible chalcogen- and halogen-related redox reactions jointly contributed to an eleven-electron transfer mode and verified the feasibility of the organic-inorganic hybrid tellurium iodide perovskite cathode. Figure 5c highlights the advantages of the (BzTEA)2TeI6 perovskite cathode that offered a record capacity compared with related references55,56,57,58,59,60,61,62,63. Long-term cycling performance was provided in Fig. 5d, e and Supplementary Fig. 29. The Zn||(BzTEA)2TeI6 battery successfully cycled 500 times at 1 A g-1 and 3 A g-1 with a high CE of approaching 98% and capacity retention of over 77% and 82%, respectively, superior to the reported counterparts (Supplementary Fig. 30). Moreover, the Zn||(BzTEA)2TeI6 battery retained 79.4% of its initial capacity after a storage period of 5 h due to the suppressed shuttle effect, while that for the Zn||“TeO2 + I2” battery was down to 16.7% (Fig. 5f and Supplementary Fig. 31). The viability of the Zn||(BzTEA)2TeI6 battery was additionally confirmed using a DC-DC converter, which maintained a consistent voltage output of 3.29 V for more than 250 min (Supplementary Fig. 32). As a summary, Fig. 5g emphasizes the importance of the special eleven-electron transfer mode actualized by the perovskite cathode and proper electrolyte design39,61,62,63,64,65,66,67,68. The structural advance endowed the zinc ion batteries with a high average voltage of 1.3 V and a large energy density of over 577 Wh kg-1Te/I. The pouch cell based on the (BzTEA)2TeI6 perovskite cathode further gave a high capacity of 113 mAh (at 10 mA cm-2) with a capacity retention of over 66% after 100 cycles (Fig. 5h). The pouch cell also presented good storage stability with a capacity loss of less than 24% after a resting time of 12 h (Supplementary Fig. 33).

Fig. 5: Cycling performance of the Zn||(BzTEA)2TeI6 battery.
figure 5

a Rate performance, b the corresponding GCD curves, and c comparison with typical cathode of zinc ion batteries. d Long-term cycling property of zinc battery at 1 A g−1, e selected GCD curves, and f the self-discharge profile. g comparison of relative aqueous zinc ion batteries regarding average voltage and energy density. h Cycling performance of the pouch cell with a high loading mass of 12 mg cm−2. The sources of referential samples in c and g are cited as ref. 55 for I2, ref. 56 for MnO2, ref. 58 for Te, ref. 59 for KMO (K0.27MnO2·0.54H2O), ref. 60 for AVO (Ag0.3V2O5), ref. 61 for IBr, ref. 63 for V2O5, ref. 62 for Co3O4, ref. 64 for PTZAN [4,4’-(10H-phenothiazine-3,7-diyl)bis(N,N-diphenylaniline)], ref. 65 for PANI (Polyaniline), ref. 66 for MnO2, ref. 67 for I2, ref. 61 for IBr, ref. 39 for Te, ref. 68 for PPy (Polypyrrole).

Full size image

Discussion

In conclusion, the designed (BzTEA)2TeI6 perovskite, as a proven methodology to actualize both B-site chalcogen and X-site halogen redox reactions, were examined upfront as the conversion-type cathode materials. The (BzTEA)2TeI6 cathode effectively confined active chalcogen and halogen elements and allowed fast transfer of chloride ions, stabilizing high-valent tellurium cations in the form of tellurium chloride ions. After coupling with the adaptive Ch0.4Zn0.6Cl1.6 · 1.5H2O electrolyte, an eleven-electron transfer was successfully realized in the Zn||(BzTEA)2TeI6 batteries as the emerging redox pairs including Cl0/Cl-1 (1.81 V), I+/I0/I- (1.64 and 1.26 V), Te6+/Te4+ (1.53 V), Te4+/Te0 (1.26 V), and Te0/Te2- (0.51 V), all of which benefited a high energy density of over 577 Wh kg-1Te/I. As a result, the Zn||(BzTEA)2TeI6 battery demonstrated a high capacity of 473 mAh g-1Te/I at 0.5 A g-1 and experienced 500 cycles at 1 A g-1 with a capacity of over 77%. The corresponding pouch cell delivered a high capacity of 113 mAh, showcasing promising storage stability.

Methods

Chemicals and reagents

Tellurium oxide (≥99%), benzyltriethylammonium chloride (BzTEACl or TEBAC, 98%), hydrogen iodide (HI, 48%), isopropanol (99%), zinc chloride (98%), choline chloride (ChCl, 98%), zinc sulfate (ZnSO4, AR), zinc acetate (Zn(OAc)2, 99%), zinc trifluoromethanesulfonate (Zn(OTF)2, 98%) and 1-methyl-2-pyrrolidinone (NMP, 98%) were purchased from Aladdin. Polyvinylidene difluoride (PVDF) was purchased from SOLVAY (Solef 1008). The conductive agent, Ketjenblack EC-300J, was purchased from Nouryon. All chemicals were used as received without further treatment.

Preparation of (BzTEA)2TeI6 microcrystal cathode and battery assembly

The perovskite microcrystals were obtained by a saturated recrystallization method. Specifically, 0.1 mmol of TeO2 and 0.2 mmol of BzTEACl were added into 1 ml HI solution. The mixture was heated to 110 °C and held for over 1 h with vigorous stirring, subsequently cooling to room temperature. The obtained black residue was washed with isopropanol and dried under a vacuum. The (BzTEA)2TeI6 perovskite microcrystals were mixed and ground with Ketjenblack and PVDF blinder in NMP with a mass ratio of 7:2:1 for 1 h. The obtained liquid slurry was evenly covered on the carbon cloth substrate, followed by a vacuum bakeout process at around 80 °C overnight. The mass loading of the perovskite cathode was estimated to be 1 ~ 1.2 mg cm-2. 39 mmol ZnCl2 was dissolved in 1 ml H2O at 100 °C and slowly cooling to 30 °C, and the clear supernatant was used as the ZnCl2 · 1.4H2O electrolyte. 0.4 mol ChCl and 0.6 mol ZnCl2 were dissolved in 1.5 mol H2O to form the Ch0.4Zn0.6Cl1.6 · 1.5H2O electrolyte. Specifically, 1 g of Ch0.4Zn0.6Cl1.6 · 1.5H2O contains 0.16 g H2O, 0.51 g ChCl, and 0.33 g ZnCl2. The cathode, zinc metal anode (100 µm, 1 cm2), and a glass fiber (420 µm, 1.1 cm2) sat in between were packed in a Swagelok cell for further evaluation (at room temperature without the need for a chamber).

Material and electrochemical characterizations

Absorption spectra were collected using a Shimadzu UV 3600 UV/visible/IR spectrophotometer, while Fourier-transform infrared (FTIR) measurements were conducted with a Perkin Elmer FT-IR spectrophotometer. Raman measurements were performed on a WITec Alpha300 R confocal Raman imaging system with a 532 nm laser. Powder X-ray diffraction (XRD) patterns were obtained using a Rigaku Smartlab X-ray diffractometer with Cu Kα radiation (λ = 1.5406 Å). X-ray photoelectron spectroscopy (XPS) was carried out on a PHI model 5802, with the carbon spectrum serving as a calibration reference. The FEI Quanta 250 scanning electron microscope (SEM) and energy dispersive spectroscopy (EDS) were utilized to examine the morphology and elemental composition of the samples. Cyclic voltammograms (CVs) and related electrochemical data were collected on a CHI 660E electrochemical workstation. The long-term stability and rate performance of the batteries were evaluated using the LAND battery testing system at room temperature (with a cut-off potential of 2.04 V).

Calculation and theoretical simulations

The b value was derived from the equation \(i=a{v}^{b}\) where \(i\) is the response current, \(v\) is the sweep rate. The b value of 0.5 signifies a purely faradaic (diffusion-controlled) process while a b value of 1 indicates a purely capacitive (surface-controlled) process, as outlined in the initial report by Lindquist et al. in 199769. The quantitative contribution of the charge storage process was evaluated by the equation \(i={k}_{1}v+{k}_{2}{v}^{0.5}\) where \({k}_{1}v\) is the surface-controlled process and \({k}_{2}{v}^{0.5}\) is the diffusion-controlled part33. The energy density was calculated by integrating the battery voltage over the specific capacity which was determined based on the mass of active elements (tellurium and iodine). The density functional theory (DFT) was employed for first-principles calculations using the Dmol3 mode within a numerical atom-centered basis function framework. The Perdew-Burke-Ernzerhof (PBE) method was used for electronic exchange-correlation interactions, while the Generalized Gradient Approximation (GGA) method with PBE formulation was applied for structural optimization70. DFT semi-core pseudopotentials were chosen for the core treatment of relativistic effects, replacing core electrons with a single effective potential71. The adsorption energy, \({E}_{{ads}}\), was calculated using the formula \({E}_{{ads}}={E}_{{ensemble}}-{E}_{{absorbent}}-{E}_{{adsorbate}}\), where the subscript represents the energy of the absorbent, adsorbate, or the entire system post adsorption72. The Raman spectrum of I5- was simulated under the excitation of 532 nm laser at 273 K, with Lorentzian smearing of 20 cm-1.