Multivalent cationic and anionic mixed redox of an Sb₂S₃ cathode toward high-capacity aluminum ion batteries

Abstract

The conventional cationic redox centers of transition-metal-based cathodes are reaching their theoretical capacity limit, which cannot match the ultra-high specific capacity contributed by the three-electron transfer reaction of an Al anode (2980 mA h g⁻¹, 8056 mA h cm⁻³), severely restraining the development of high-energy rechargeable Al-ion batteries (RAIBs). In this work, we propose the multivalent cationic and anionic mixed redox chemistry of Sb₂S₃ as a promising way out of this problem. The energy storage is induced by the cumulative Sb-related cationic (Sb(+3) ⇔ Sb(+5)) and S-related anionic (S(−2) ⇔ S(0)) mixed 10-electron transfer reversible redox reaction during the charge/discharge process, which has been elucidated here by extensive electrochemical measurements and characterizations. Furthermore, excellent electrochemical performances are realized due to the carbon-based interlayer effectively blocking the charging products of the Sb-based cationic cluster (SbCl₄⁺), by a dual defense mechanism that integrates the physical barrier of the porous structure and the powerful chemical adsorption ability of the oxygen groups. The Sb₂S₃ cathode could deliver a discharge specific capacity of 756 mA h g⁻¹ at 100 mA g⁻¹. The finding is that the joint multivalent cationic and anionic redox chemistry proposed in this work opens up new opportunities for designing high-performance electrodes for advanced rechargeable batteries.

---

Introduction

The high reserves of the metal Al, and its high safety, low cost, and non-toxicity have made the aluminum-ion battery (AIB) a competitive metal-ion battery system. Significantly, the reversible three-electron transfer reaction of the Al anode can deliver an ultra-high specific capacity of 2980 mA h g⁻¹ and 8056 mA h cm⁻³,¹ promoting AIBs as a primary candidate for the next-generation energy storage system. Nonetheless, the energy density and sustainability of AIBs until now are still not comparable to Li-ion batteries, with the lack of suitable multi-electron-based cathodes, to match the highly reversible three-electron transfer reaction of the Al anode, being a major obstacle. Therefore, an intense research effort has been spurred into action for the discovery of novel cathode materials for AIBs with a multi-electron transfer reaction, large capacity, and long cycle life.

Among the cathode candidates for AIBs considered at present, transition-metal-sulfide (TMS) cathodes are the most attractive, owing to their multi-electron transfer reactions.² But, since the advent of AIBs, the multi-electron transfer reaction of these TMS has been considered/proved exclusively based on cationic redox electrochemistry. Unfortunately, without the contribution of S-based anionic redox, the average number of electrons transferred per element based solely on the cationic redox of these cathodes is relatively small,^3–5 thus leading to insufficient theoretical capacity for real-world implementation. Therefore, introducing a multivalent cationic and S-based anionic mixed redox is an effective approach to boost the specific capacity and energy density via accessing the charge transfer from all the elements of TMS, but it also faces huge challenges.

Here, we report a novel one-dimensional nanorod-like Sb₂S₃ material as a cathode for high-capacity RAIBs for the first time. Unlike previously reported TMS cathodes, where energy storage is induced by the redox reaction of the cations, the mechanism from which the high capacity is achieved lies in the reversible redox of the S-related anions in addition to the metal Sb-related cations of Sb₂S₃. Based on the cumulative Sb-related cationic (Sb(+3) ⇔ Sb(+5)) and S-related anionic (S(−2) ⇔ S(0)) mixed reversible redox processes, the Sb₂S₃ cathode can deliver a mixed 10-electron transfer reaction, holding a surprising theoretical capacity of 789 mA h g⁻¹. Furthermore, a graphene aerogel interlayer as a preventive measure for tackling the possible shuttle effect of the soluble ion-based products was proved to be a sensible option. Density functional theory (DFT) calculations showed that reduced graphene oxide with oxygen-containing functional groups could powerfully chemically adsorb the cationic clusters of SbCl₄⁺. As a result, this novel Sb₂S₃ cathode delivers an initial discharge specific capacity of 756 mA h g⁻¹ at 100 mA g⁻¹, nearing its theoretical limit. Even at a high current density of 1000 mA g⁻¹, the discharge specific capacity is as high as ca. 500 mA h g⁻¹, remaining at 363 mA h g⁻¹ after 220 cycles. Considering both the discharge capacity and rate performance, the cathode prepared herein surpasses all the transition-metal-based cathode materials reported to date.

Results and discussion

Preparation and morphology of Sb₂S₃ nanorods

The synthesis process of Sb₂S₃ nanorods is shown in Fig. 1a, and detailed experiments can be seen in the ESI.† Antimony trichloride was hydrolyzed to produce antimony oxychloride, which was vulcanized by a solvothermal reaction to produce rod-shaped Sb₂S₃. Fig. 1b shows the XRD pattern of the synthesized Sb₂S₃. All the characteristic peaks match the orthorhombic Sb₂S₃ phase (JCPDS card no. 00-006-0474). The morphology of Sb₂S₃ was characterized by SEM and TEM (Fig. 1c–e). Fig. 1c shows that the synthesized Sb₂S₃ nanorods are aggregated from multiple thinner nanorods with a diameter of ca. 100 nm. The TEM image clearly shows that the structure of Sb₂S₃ is very uniform. It can be seen from the electron diffraction pattern of the selected area (Fig. 1e) that the synthesized Sb₂S₃ is in a single crystal state, and the lattice spacing of the (110) crystal plane is 0.8 nm.⁶

Fig. 1 (a) Schematic diagram of the preparation process of Sb₂S₃ nanorods. (b) XRD pattern; (c) SEM; (d) TEM image; (e) HR-TEM image and the selected-area electron diffraction pattern of Sb₂S₃.

Electrochemical performance of the Sb₂S₃ cathode

Sb₂S₃ with a graphene aerogel interlayer was assembled in a Swagelok-type battery (Fig. S1†) to evaluate the electrochemical performance of Sb₂S₃ as an RAIB battery cathode material. As shown in Fig. 2a, the CV curve of the Sb₂S₃ cathode has two obvious oxidation peaks in the voltage range of 0.8 to 1.9 V and the corresponding reduction peaks appear in the voltage range of 0.9 to 0.3 V. In order to further understand the kinetics of the Sb₂S₃ cathode, the relation between the peak current and scan rate can be expressed as i = av^b, where i is the response current (mA) and v is the scan rate (mV s⁻¹); a and b are adjustable values. A b value of 0.5 generally indicates a diffusion-controlled process, while a value of 1.0 suggests that the reaction is a surface charge transfer process.^7,8 By fitting the redox peaks at different scan rates it was found that the slopes of the fitting curves for different redox peaks were 0.96, 0.92, 0.90, and 0.95, respectively, indicating an interplay between surface- and diffusion-controlled reactions, but predominantly a surface-driven pseudo-capacitive action (Fig. 2c). The calculated contribution of pseudo-capacitance and ion diffusion to battery capacity was calculated according to the formula i(V) = k₁v + k₂v^1/2, where i(V) is the total current response at a given potential V, and k₁ and k₂ are the capacitive and diffusion contribution, respectively. As shown in Fig. 2b, the pseudo-capacitance contribution to the total charge storage area curves at a scan rate of 1.0 mV s⁻¹ was 94.9%, indicating that the Sb₂S₃ cathode exhibits almost complete pseudo-capacitive properties.^9,10

Fig. 2 Electrochemical performance of Sb₂S₃: (a) CV curves at different scan rates. (b) Comparison graph of pseudo-capacitance contribution and CV curve at a scan rate of 1.0 mV s⁻¹. (c) The log(i)–log(v) plots of peaks and their linear fitting results. (d) The galvanostatic charge/discharge curve at 1000 mA g⁻¹. (e) The rate performance at different current densities. Cycle performance curves at (f) 100 mA g⁻¹ and (g) 1000 mA g⁻¹. Notes: in (e) and (f) share the same y-axis of coulombic efficiency.

Fig. 2d shows the galvanostatic charge/discharge curve at a high current density of 1000 mA g⁻¹. The assembled battery was discharged first, and the initial discharge capacity was 75 mA h g⁻¹. After nearly 20 cycles of activation, the charge–discharge specific capacity reached a stable level of 489 mA h g⁻¹. Two charging plateaus at ca. 1.15 and 1.60 V, and the corresponding discharge plateaus at ca. 0.65 and 0.80 V were observed, in full agreement with the CV curve. While, at a small current density of 100 mA g⁻¹, we compared two charging protocols: charge-first, and discharge-first (Fig. S2†). The initial discharge capacity of discharge-first was small, only 140 mA h g⁻¹, which is mainly contributed by the formation of an SEI film, GA interlayer and conductive carbon black, but the following charging curve of the two charging protocols showed almost no difference in capacity or voltage plateau. After one cycle of activation, the capacity and voltage plateau of the charge–discharge curve tended to be stable, and the discharge specific capacity could reach 756 mA h g⁻¹ with a cell-level energy density of 80 W h kg⁻¹ (Fig. S3a†).^11,12 Furthermore, on the charging curve at this current density, a third-stage charging plateau appeared in the voltage range 1.8–2.0 V, which can be judged to be the oxidation of S (S⁰ → S^x+).¹¹ However, no corresponding high-potential discharge plateau was observed on the discharge curve, which resulted in a low coulombic efficiency (∼95%). To find a more suitable voltage window, cycling performance tests were carried out in different voltage ranges at 1000 mA g⁻¹ (Fig. S4 and S5†). In the voltage range 0.2–1.8 V, the capacity stabilized at about 225 mA h g⁻¹ after 50 cycles. After increasing the cut-off voltage of charging to 2.0 V, the capacity rapidly increased to about 500 mA h g⁻¹, which is due to the increasing cut-off voltage being able to make the reaction more sufficient. The discharge potentials were adjusted to 0.1 V and 0.01 V, respectively, and the capacities were comparable, but the cycle decay rate of the battery was lower in the working region of 0.01–2.0 V. The capacity was 318 mA h g⁻¹ after 120 cycles in the working range 0.2–2.0 V. Fig. 2e shows the rate performance curves at different current densities. The discharge specific capacities were 660, 630, 604, and 450 mA h g⁻¹ at current densities of 200, 300, 500, and 2000 mA g⁻¹, respectively. After gradually reducing the current density to 100 mA g⁻¹, the specific discharge capacity could be maintained at 615 mA h g⁻¹. At a large current density of 1000 mA g⁻¹, the capacity of the battery required several cycles of activation to be fully developed, and was maintained at 363 mA h g⁻¹ after 220 cycles with coulombic efficiencies much closer to 100% (Fig. 2g). Fig. S3c† demonstrates the cycling performance of the Sb₂S₃ electrode with higher loadings. The capacities remained at 311, 260 and 175 mA h g⁻¹ after 200 cycles at high loadings of 2.6, 4.5 and 7.5 mg, respectively. Compared to reported transition-metal-based cathodes (Table S1†), the discharge capacity and the rate performance of Sb₂S₃ are among those of the best cathode materials reported for AIBs. It is remarkable, however, that the reversible charge specific capacity of the Sb₂S₃ cathode (756 mA h g⁻¹) is much higher than the theoretical specific capacity (473 mA h g⁻¹) of the only Sb-based cationic redox reaction from Sb(+3) to Sb(0). In view of the above analysis of electrochemical properties, a bold hypothesis emerges naturally that the S-based anionic oxidation reaction also contributes reversible charge specific capacity to the Sb₂S₃ cathode during the charging process. Furthermore, similar cathodic peaks in the range of 1.4 to 1.9 V (Fig. S6†) were observed by comparing the CV curves of Sb₂S₃ and elemental sulfur (oxidized from S(−2) to S(0)),^13,14 which further echoed our hypothesis of the S-based anionic reaction in the Sb₂S₃ cathode. This navigated us to pursue the possible energy storage mechanism of the cationic and anionic mixed redox reaction in the Sb₂S₃ cathode, which seemed promising.

Characterization of energy storage mechanism

To explore the possible cationic and anionic mixed redox based energy storage mechanism of the Sb₂S₃ cathode in RAIBs, TEM, in situ XRD, ex situ XPS, and ex situ Raman measurements were conducted herein to identify the electrode components and structures in different charge/discharge stages. Comparing the TEM image and the corresponding element mapping images of the Sb₂S₃ cathode in the initial state and the fully charged/discharged states (Fig. S7 and S8†), it can be shown that the Sb₂S₃ cathode underwent a phase change reaction during the charging process. To prove that, in situ XRD (Fig. S9†) was conducted to monitor the structural changes of the Sb₂S₃ cathode during the charging and discharging process. Compared with the XRD pattern of pure Sb₂S₃ powder shown in Fig. 1b, the relative strength of the characteristic peak for the Sb₂S₃ cathode, collected by an X-ray penetrating Be current collector, has changed to a certain extent, but there is no change in the peak position. Fig. S10† shows that the peaks of the Be current collector exist at the angles of 45.81, 50.93, 52.79, and 70.89°, and the peak intensity is much higher than that of the electrode material.¹⁵ In order to improve the sensitivity of in situ XRD detection, the range of signal acquisition was set at 5–50°. Fig. 3a shows the structural changes of the Sb₂S₃ during the first discharge and the second complete charge/discharge cycle. The main characteristic peaks of Sb₂S₃ exist at 22.30, 25.04, 28.50, 29.26, and 35.52°. During the initial discharge process, the intensity of the monitored Sb₂S₃ peak did not change, and no new peaks appeared, which indicated that Sb₂S₃ was not reduced into crystal Sb during the first discharge process.¹⁶ Then, during the charging process, the intensity of the characteristic peak of Sb₂S₃ gradually weakened, which indicates that the Sb₂S₃ cathode was consumed by a phase change reaction. In addition, no deviation of the characteristic peaks of the cathode was observed, and no new peaks appeared, so it can be speculated that the energy storage mechanism is different from the conventional intercalation conversion mechanism.^3,17–19 During the discharge process, the intensity of these characteristic peaks for Sb₂S₃ gradually returned to the initial state completely, and no new peaks were detected, indicating that the charging product could be reduced sufficiently thus leading to the regeneration of crystal Sb₂S₃. This fully demonstrates the reversibility of the battery reaction.

Fig. 3 (a) In situ XRD patterns of Sb₂S₃ electrodes in different charge/discharge states. (b and c) Ex situ XPS spectra of Sb 3d and S 2p of Sb₂S₃ electrodes in different charge/discharge states. (d) Ex situ Raman spectra of Sb₂S₃ electrodes in different charge/discharge states. (e) Raman spectra of the electrolyte after fully charged sate, pure electrolyte, and electrolyte with SbCl₅.

Fig. 3b and c show the ex situ XPS spectra of Sb 3d and S 2p of the Sb₂S₃ electrodes in different charge/discharge states. As shown in Fig. 3b, the Sb spectrum of pristine Sb₂S₃ has two characteristic peaks at 529.48 and 538.78 eV, which correspond to the binding energy of Sb(+3) 3d_3/2 and 3d_5/2.²⁰ For the S spectrum of pristine Sb₂S₃ (Fig. 3c), a pair of peaks located at 160.78 eV (2p_3/2) and 161.88 eV (2p_1/2) confirmed the existence of S(−2).^21,22 By charging to 1.2 V, the Sb 3d_3/2 and 3d_5/2 spectra totally shifted to higher binding energies of 531.98 and 539.98 eV, respectively, corresponding to Sb(+5) (Fig. S11a†).^20,23,24 This indicated that the Sb(+3) was completely oxidized to a higher valence of Sb(+5). However, a new pair of peaks located at 163.58 eV (2p_3/2) and 164.58 eV (2p_1/2) was observed in the S spectrum at the charging voltage of 1.2 V that confirmed the existence of elemental S. When fully charged to 2.0 V, the same characteristic peaks of Sb(+5) were still observed in the Sb spectrum, indicating that Sb(+5)-based charging products could exist stably under this voltage in RAIBs. While for the S spectrum in the same fully charged state, the S 2p peak totally shifted to a higher binding energy of 163.58 eV (2p_3/2) and 164.58 eV (2p_1/2), indicating that S(−2) was well oxidized to S(0) (Fig. S11b†).^25,26 Based on the research into Li–S²⁷ and Al–S batteries,²⁸ S is difficult to detect in the form of the crystalline state after cycling owing to the regenerated S existing in the form of amorphous S,^29,30 consistent with the results from in situ XRD where no crystal S was observed. In terms of the above analysis, the reaction mechanism during the charging process of the Sb₂S₃ cathode could be divided into two stages: namely, the Sb-based oxidation process, then the S-based oxidation process. During the discharge progress, in addition to the peaks of Sb(+5) 3d_3/2 and 3d_5/2, the peaks of Sb(+3) 3d_3/2 and 3d_5/2 reappeared at the discharge voltage of 0.7 V, indicating the reduction reaction from Sb(+5) to Sb(+3). No change was observed in the S spectrum, the S(0) 2p_1/2 and 2p_3/2 peaks totally remained at the discharging voltage of 0.7 V. Subsequently, by discharging to 0.01 V, the intensity of Sb(+3) peaks at 529.48 and 538.78 eV increased significantly, indicating that Sb(+5) was reversibly reduced to Sb(+3). While for the S spectrum in the same fully discharged state, the S(−2) 2p peaks reappeared at 160.78 and 161.88 eV, which were derived from Sb₂S₃, indicating that S(0) was also reduced to S(−2). In order to reduce the influence of residual electrolyte on the XPS test results, we performed argon ion etching (depth of 5 nm) on electrodes in different charge and discharge states to expose the electrode materials adhered to by the electrolyte (Fig. S12†), obtaining results completely consistent with Fig. 3b and c, so the removal of the surface electrolyte allows the electrode in the intermediate states to present more abundant information. Based on the ex situ XPS results, the energy storage mechanism of Sb₂S₃ could be preliminarily identified as the cumulative Sb-related cationic (Sb(+3) ⇔ Sb(+5)) and S-related anionic (S(−2) ⇔ S(0)) mixed reversible redox processes, and it can be determined that peaks I and III in Fig. S13† are redox peak pairs (1.19 V/0.69 V) of Sb, while peak II and peak IV are S-related redox peaks (1.63 V/0.56 V).

To further identify the Sb(+5)-based and S(0)-based charging products, ex situ Raman spectroscopies were conducted herein. As shown in Fig. 3d, pristine Sb₂S₃ had obvious characteristic peaks at 146.1, 188.3, 251.5, 297.5, and 449.2 cm⁻¹.^31,32 When charged to 2.0 V, these peaks of Sb₂S₃ disappeared, and three new peaks at 152.7, 220.8, and 471.3 cm⁻¹, which were completely derived from the peaks of elemental S,³³ were observed, consistent with the results from ex situ XPS. During the discharge process, the characteristic peak of S almost completely disappeared, and the peak representing Sb₂S₃ in the range 250–330 cm⁻¹ reappeared. The aforementioned results are further good proof of the S-related anionic (S(−2) ⇔ S(0)) reversible redox reaction. As we revealed in our previous research work on Al–Se batteries,³⁴ a high-valent charging product such as Se(+4) is easily combined with Cl⁻, forming chloro-complexed cations such as SeCl₃⁺, in Lewis-acid chloroaluminate melts ionic liquid electrolytes. So, the pentavalent Sb(+5) are most likely to form Sb(+5)-based chloro-complexed cations of SbCl₄⁺ in the Lewis-acid chloroaluminate melts ionic liquid electrolytes.^35,36 To verify this hypothesis, pure antimony pentachloride (SbCl₅) (Fig. S14†) and SbCl₅/AlCl₃/[EMIM]Cl melts ionic liquid were synthesized. Then, Raman spectroscopy was conducted herein to identify the coordination structure for the Sb(+5)-based charge products. For the pure SbCl₅, as shown in Fig. S15a,† characteristic peaks at 178.6, 304.3, 354.8, and 395.3 cm⁻¹ were observed in its Raman spectrum.³⁷ According to previous studies, SbCl₅ can interact with chloroaluminate, such as Al₂Cl₇⁻ to generate chloro-complexed cations of SbCl₄⁺ when SbCl₅ is dissolved in chloroaluminate melts,³⁸ showing characteristic peaks at 127.3, 178.6, 281.0, 330.1 and 491.0 cm⁻¹ in the Raman spectrum (Fig. S15b†). In addition, electrolyte trapped within a GA interlayer near the cathode side from a fully charged battery was subjected to a Raman test to observe the change in electrolyte composition. As shown in Fig. 3e, in addition to the peaks of AlCl₄⁻ (348.2 cm⁻¹), Al₂Cl₇⁻ (435 cm⁻¹) and EMIM⁺ (596.8 cm⁻¹),³⁹ new peaks at 178.6, 281.0, and 330.1 cm⁻¹ appeared in the Raman spectrum, consistent with the characteristic peaks of SbCl₄⁺, indicating that the Sb(+3) in Sb₂S₃ was oxidized to SbCl₄⁺ during the charging process. Combining the reported literature (Table S2†) and the above measurements, the energy storage mechanism for the Sb-related cationic and S-related anionic mixed reversible redox reaction of the Sb₂S₃ cathode in RAIBs can be summarized as follows:

Cathode reaction:

Sb₂S₃ + 16AlCl₄⁻ ↔ 2SbCl₄⁺ + 3S + 8Al₂Cl₇⁻ + 10e⁻

Anode reaction:

Total reaction:

Accordingly, a mixed 10-electron transfer reaction (Fig. 4a and S16†) is realized by the joint cationic and anionic reversible redox of Sb₂S₃, holding a surprising theoretical capacity of 789 mA h g⁻¹. The formation and conversion of elemental S contributed to the higher theoretical capacity of the battery and explained the large polarization on the charge–discharge curves in Fig. 2d, since sulfur is an electronic insulator and has slower reaction kinetics, which also leads to lower energy efficiency (52%).^40,41 It is noteworthy that the soluble charging products of SbCl₄⁺ could shuttle into the ionic liquid electrolyte, resulting in the capacity decay of the battery. Inspired by the Li–S battery, a carbon-based interlayer was used here to mitigate the shuttle effect of the soluble products. As expected, improved electrochemical performances were achieved here by using a GA interlayer.

Fig. 4 Schematic diagram of the energy storage mechanism of the Sb₂S₃ cathode materials during the charging process (a). Theoretical calculations of adsorption energies and structures of SbCl₄⁺ adsorbed on S, Sb₂S₃ and rGO (b).

Contribution of the GA interlayer to battery performance

The capacity contributions of the GA interlayer were investigated, as presented in Fig. S17.† The discharge specific capacity of pure GA presents only 25 mA h g⁻¹ at a current density of 100 mA g⁻¹. Fig. S18† shows the galvanostatic charge/discharge curves and cycle performance curves of the Sb₂S₃ cathode without a GA interlayer. It is found that the battery without a GA interlayer has similar charging and discharging platforms to the battery using a GA interlayer, but the initial discharge specific capacity presents only 191 mA h g⁻¹, which is far less than its theoretical specific capacity. Even worse, the discharge specific capacity reduced dramatically to 37 mA h g⁻¹ after 50 cycles. As revealed, the poor cycling performance was caused by the shuttle effect of SbCl₄⁺ during the repeated charge/discharge processes, which could be effectively improved by using a GA interlayer.⁴² Fig. S19† shows the SEM morphology of the GA interlayer with a three-dimensional porous structure. The batteries after long cycles were disassembled to observe the changes in the morphology and structure of the Sb₂S₃ electrode and GA interlayer. Sb₂S₃ nanorods can be clearly seen on the initial electrode, and the elemental contents of S and Sb are close to the theoretical molar ratio of 1.5 (Fig. S20†). After 50 cycles, the rod-shaped Sb₂S₃ showed some defects and had decreased, while after 200 cycles, almost no Sb₂S₃ could be seen on the electrode surface, all of which had been converted into smaller nanoparticles, and the content of S and Sb was also significantly reduced. In contrast, a large amount of uniformly distributed Sb and S was observed on the GA interlayer after long cycling, which indicated that the original rod-like Sb₂S₃ had transformed and recombined, and the GA interlayer had effectively adsorbed and blocked the shuttling of charging products (Fig. S21†). As shown in Fig. S22,† the FTIR and XPS spectra illustrate the abundant oxygen-containing functional groups of –OH, –CO and –COOH in the GA interlayer. To further understand the mechanism whereby the GA interlayer inhibits the shuttle effect of SbCl₄⁺, density functional theory (DFT) calculations were utilized to explore the possible enhanced chemisorption ability of a carbon-based interlayer toward soluble SbCl₄⁺ (Fig. 4b). According to the DFT calculations, the binding energies for S and SbCl₄⁺, Sb₂S₃ and SbCl₄⁺, reduced graphene oxide (rGO) and SbCl₄⁺ were estimated to be −1.70, −4.12 and −5.67 eV, respectively. Owing to the low chemisorption between the accumulated S and the soluble SbCl₄⁺ during the cycling process, accordingly, severe shuttle effects lead to capacity decay. Fortunately, the DFT results indicate that reduced graphene oxide with oxygen-containing functional groups, such as –OH, –CO and –COOH, exhibited satisfactory chemical adsorption ability for SbCl₄⁺. So, a GA interlayer, synthesized from graphene oxide with a gradient content of oxygen-containing functional groups (Fig. S21b†), was introduced into the battery and effectively suppressed the shuttle effect of SbCl₄⁺ by a dual defense mechanism that integrated the physical barrier of a porous structure and the powerful chemical adsorption ability of the oxygen groups. As a result, the battery exhibits a superior discharge capacity close to its theoretical capacity and a stable cycle performance. Furthermore, Fig. S23† demonstrates the self-discharge performance of the Al–Sb₂S₃ battery using a GA interlayer. When the resting time is 1 or 2 h, the discharge capacities show almost no attenuation, about 575 mA h g⁻¹, and the coulombic efficiencies are close to 100%. When the resting time increases to 6 h, the discharge capacity decays weakly and remains at 554 mA h g⁻¹. However, when the standing time continued to increase to 12, 24, and 48 h, the capacity decays are more obvious, which are maintained at 504, 420, and 350 mA h g⁻¹, respectively. This indicates that the GA interlayer has indeed achieved superior results in suppressing the shuttle effect of charging products in a short time, but more efforts are needed to fundamentally solve the problem, thereby further improving the battery's stability.

Conclusions

In summary, unlike previously reported TMS cathode materials where energy storage is induced by the redox reaction of the cations in it alone, multivalent cationic and anionic mixed redox reactions which are imperative to further improve the battery performance were introduced into RAIBs by using a novel Sb₂S₃ cathode. The energy storage properties of the Sb₂S₃ cathode were understood to follow a joint Sb-related cationic (Sb(+3) ⇔ Sb(+5)) and S-related anionic (S(−2) ⇔ S(0)) mixed reversible redox mechanism. DFT calculations confirm that the strong chemisorption between the carbon-based interlayer with oxygen-containing functional groups and SbCl₄⁺, effectively suppressed the shuttle effects. Furthermore, the Sb₂S₃ cathode exhibited outstanding cycling performance of 363 mA h g⁻¹ at 1000 mA g⁻¹ after 220 cycles with a record-setting cyclic performance achieved in the documented AIBs among all the metal sulfide cathode materials reported to date. The discovery of multivalent cationic and anionic mixed redox activity in metal sulfides might be a feasible solution to the pursuit of cathode materials for the further development of RAIBs.

Author contributions

T. Li, T. Cai: conceptualization, methodology, validation, writing – original draft, writing – review & editing, visualization; D. Wang, H. Hu: visualization; Y. Zhang, Y. Cui, X. Li, W. Xing: validation, investigation; L. Zhao: software, formal analysis; Z. Yan: writing – review & editing, supervision, funding acquisition, project administration. All authors have given approval to the final version of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the Taishan Scholar Foundation (tspd20210308), the Key Projects of China National Key R&D Plan (2018YFE0118200) and the Key Projects of Shandong Key R&D plan (2019JZZY010506), the 111 Progam of National College Disciplinary Innovation (B03031), the National Natural Science Foundation of China (51877216, 21905300, 22109180), the Natural Science Foundation of Shandong Province (ZR2020MB078, ZR202103040491), the Fundamental Research Funds for the Central Universities (21CX06011A), the Major Program of Shandong Province Natural Science Foundation (ZR201801280009), the PetroChina Innovation Foundation (2018D-5007-0504), the Fundamental Research Funds for the Central Universities (20CX060103A), the Postdoctoral Applied Research Program of Qingdao (qdyy20200077), and the Innovation Project for Postgraduates of China University of Petroleum (East China) (YCX2021070).

References

1. H. Yang, H. Li, J. Li, Z. Sun, K. He, H. M. Cheng and F. Li, Angew. Chem., Int. Ed. Engl., 2019, 58, 11978–11996.
2. J. Tu, W. L. Song., H. Lei, Z. Yu, L. L. Chen, M. Wang and S. Jiao, Chem. Rev., 2021, 121, 4903–4961.
3. R. Zhuang, Z. Huang, S. Wang, J. Qiao, J.-C. Wu and J. Yang, Chem. Eng. J., 2021, 409, 128235.
4. L. Xing, K. A. Owusu, X. Liu, J. Meng, K. Wang, Q. An and L. Mai, Nano Energy, 2021, 79, 105384.
5. Y. Hu, H. Huang, D. Yu, X. Wang, L. Li, H. Hu, X. Zhu, S. Peng and L. Wang, Nanomicro Lett., 2021, 13, 159.
6. J. Varghese, S. Barth, L. Keeney, R. W. Whatmore and J. D. Holmes, Nano Lett., 2012, 12, 868–872.
7. J. Liu, J. Wang, C. Xu, H. Jiang, C. Li, L. Zhang, J. Lin and Z. X. Shen, Adv. Sci., 2018, 5, 1700322.
8. S. Fleischmann, J. B. Mitchell, R. Wang, C. Zhan, D. E. Jiang, V. Presser and V. Augustyn, Chem. Rev., 2020, 120, 6738–6782.
9. S. Yao, J. Cui, Y. Deng, W. G. Chong, J. Wu, M. Ihsan-Ul-Haq, Y.-W. Mai and J.-K. Kim, Energy Storage Mater., 2019, 20, 36–45.
10. W. Zhan, M. Zhu, J. Lan, H. Wang, H. Yuan, X. Yang and G. Sui, Chem. Eng. J., 2021, 408, 128007.
11. H. Li, R. Meng, Y. Guo, B. Chen, Y. Jiao, C. Ye, Y. Long, A. Tadich, Q. H. Yang, M. Jaroniec and S. Z. Qiao, Nat. Commun., 2021, 12, 5714.
12. T. Zhang, T. Cai, W. Xing, T. Li, B. Liang, H. Hu, L. Zhao, X. Li and Z. Yan, Energy Storage Mater., 2021, 41, 667–676.
13. D. Zhang, X. Zhang, B. Wang, S. He, S. Liu, M. Tang and H. Yu, J. Mater. Chem. A, 2021, 9, 8966–8974.
14. T. Gao, X. Li, X. Wang, J. Hu, F. Han, X. Fan, L. Suo, A. J. Pearse, S. B. Lee, G. W. Rubloff, K. J. Gaskell, M. Noked and C. Wang, Angew. Chem., Int. Ed. Engl., 2016, 55, 9898–9901.
15. L. D. Ellis, B. N. Wilkes, T. D. Hatchard and M. N. Obrovac, J. Electrochem. Soc., 2014, 161, A416–A421.
16. W. Guan, L. Wang, J. Tu and S. Jiao, J. Electrochem. Soc., 2020, 167, 080541.
17. Y. Tong, A. Gao, Q. Zhang, T. Gao, J. Yue, F. Meng, Y. Gong, S. Xi, Z. Lin, M. Mao, S. Peng, X. Wang, D. Xiao, D. Su, Y. Luo, H. Li, L. Chen, L. Suo and L. Gu, Energy Storage Mater., 2021, 37, 87–93.
18. Z. Zhao, Z. Hu, Q. Li, H. Li, X. Zhang, Y. Zhuang, F. Wang and G. Yu, Nano Today, 2020, 32, 100870.
19. K. Liang, L. Ju, S. Koul, A. Kushima and Y. Yang, Adv. Energy Mater., 2019, 9, 1802543.
20. T. Birchall, J. A. Connor and I. H. Hillier, J. Chem. Soc. Dalton Trans, 1975, 2003–2006.
21. V. P. Zakaznova-Herzog, S. L. Harmer, H. W. Nesbitt, G. M. Bancroft, R. Flemming and A. R. Pratt, Surf. Sci., 2006, 600, 348–356.
22. J. Grigas, E. Talik and V. Lazauskas, Phase Transitions, 2002, 75, 323–337.
23. J. Wang, D. Feng, W. Wu, M. Zeng and Y. Li, Polym. Degrad. Stab., 1991, 31, 129–140.
24. R. Delobel, H. Baussart, J. Leroy, J. Grimblot and L. Gengembre, J. Chem. Soc., Faraday Trans., 1983, 79, 879–891.
25. N. Wang, Y. Wang, Z. Bai, Z. Fang, X. Zhang, Z. Xu, Y. Ding, X. Xu, Y. Du, S. Dou and G. Yu, Energy Environ. Sci., 2020, 13, 562–570.
26. K. Zhang, T. H. Lee, J. H. Cha, R. S. Varma, J. W. Choi, H. W. Jang and M. Shokouhimehr, Sci. Rep., 2019, 9, 13573.
27. C. Barchasz, F. Molton, C. Duboc, J. C. Lepretre, S. Patoux and F. Alloin, Anal. Chem., 2012, 84, 3973–3980.
28. X. Zheng, R. Tang, Y. Zhang, L. Ma, X. Wang, Y. Dong, G. Kong and L. Wei, Sustainable Energy Fuels, 2020, 4, 1630–1641.
29. X. Yu and A. Manthiram, Adv. Energy Mater., 2017, 7, 1700561.
30. Y. Guo, H. Jin, Z. Qi, Z. Hu, H. Ji and L.-J. Wan, Adv. Funct. Mater., 2019, 29, 1807676.
31. Y. A. Sorb, V. Rajaji, P. S. Malavi, U. Subbarao, P. Halappa, S. C. Peter, S. Karmakar and C. Narayana, J. Phys.: Condens. Matter, 2016, 28, 015602.
32. P. Makreski, G. Petruševski, S. Ugarković and G. Jovanovski, Vib. Spectrosc., 2013, 68, 177–182.
33. A. T. Ward, J. Phys. Chem., 1968, 72, 4133–4139.
34. T. Zhang, T. Cai, W. Xing, T. Li, B. Liang, H. Hu, L. Zhao, X. Li and Z. Yan, Energy Storage Mater., 2021, 41, 667–676.
35. Z. Li, J. Liu, X. Huo, J. Li and F. Kang, ACS Appl. Mater. Interfaces, 2019, 11, 45709–45716.
36. H. Jiao, D. Tian, S. Li, C. Fu and S. Jiao, ACS Appl. Energy Mater., 2018, 1, 4924–4930.
37. H. A. Szymanski, R. Yelin and L. Marabella, J. Chem. Phys., 1967, 47, 1877–1879.
38. J. W. J. Casteel, P. Kolb, N. LeBlond, P. Hélène, A. Mercier and G. J. Schrobilgen, Inorg. Chem., 1996, 35, 929–942.
39. H. Sun, G. Zhu, X. Xu, M. Liao, Y. Y. Li, M. Angell, M. Gu, Y. Zhu, W. H. Hung, J. Li, Y. Kuang, Y. Meng, M. C. Lin, H. Peng and H. Dai, Nat. Commun., 2019, 10, 3302.
40. D. Zhang, X. Zhang, B. Wang, S. He, S. Liu, M. Tang and H. Yu, J. Mater. Chem. A, 2021, 9, 8966–8974.
41. Y. Guo, Z. Hu, J. Wang, Z. Peng, J. Zhu, H. Ji and L. J. Wan, Angew. Chem., Int. Ed. Engl., 2020, 59, 22963–22967.
42. X. Zhang, S. Jiao, J. Tu, W.-L. Song, X. Xiao, S. Li, M. Wang, H. Lei, D. Tian, H. Chen and D. Fang, Energy Environ. Sci., 2019, 12, 1918–1927.

---

Footnotes

† Electronic supplementary information (ESI) available. See https://doi.org/10.1039/d2ta02049a

‡ These authors contributed equally to this work.

---