1 Introduction

Up to now, many remote areas still do not have developed power grid or access to environmentally sound and affordable energy [1, 2]. Therefore, the development of self-charging power systems is very significant to solve the problem of energy poverty in underdeveloped areas. To solve this problem, the integrated systems have been developed, combining rechargeable battery (e.g., alkali batteries, zinc ion batteries (ZIBs)) with external energy collection systems (e.g., solar cells, thermobatteries, nanogenerators) [3,4,5,6,7,8]. However, under certain conditions, external energy collection systems are not always sustainable access, for example, solar cells need sunlight, thermobatteries need temperature difference and nanogenerators need additional mechanical energy [9,10,11]. In addition, the complex and large component of the integrated self-charging power systems is not suitable for the rapid development of portable and wearable electronics. Therefore, it is necessary to find a new type of self-charging power systems with simplified configuration and suitable for various environments.

Oxygen is known to be a rich source of energy from air and can be used to convert chemical energy into electrical energy through redox reactions. Therefore, combining the chemical energy of oxygen with the electrochemical energy of electrode material is an effective method to obtain self-charging batteries. Considering the instability of alkali batteries in the air [12,13,14,15,16], aqueous batteries seem to be the only option for self-charging batteries. In particular, ZIBs have recently stood out because the Zn anode possesses inherent merits such as high capacity (820 mAh g–1), low redox potential (− 0.76 V vs. SHE), high safety, environmental benignity and low cost [17,18,19,20,21,22,23].

At present, some researches have been made in the combination of ZIBs and oxygen to produce air-rechargeable ZIBs. For example, Chen et al. at first reported a air-rechargeable ZIBs based on GaVO cathodes, which can be charged by oxygen for 36 h to reach 1.05 V open-circuit voltage, providing a capacity of 239 mAh g−1 at 0.1 A g−1 [24]. However, the air-rechargeable ZIBs show limited cycle life (only five cycles), so the exploration of ZIBs with higher life was the next logical step. Subsequently, Wang et al. reported a air-rechargeable ZIBs based on poly(1,5-naphthaleneamine) cathode and alkaline electrolyte. Benefited from the weak coordination reaction between the cathode and cation, the cycle life of air-rechargeable ZIBs is improved (100 cycles). However, its discharge capacity is only 186 mAh g−1 at 0.2 A g−1 [25]. These works undoubtedly provide an effective strategy for the development of air-rechargeable ZIBs. However, it is still necessary to develop new air-rechargeable cathodes to obtain air-rechargeable ZIBs with both long life and high capacity.

As well known, among many ZIBs cathodes materials, organic–inorganic hybrid materials (OIHMs) have long cycle life and high capacity, benefiting from that organics can manage to enhance the electronic transfer, suppress the cathode dissolution and moderate the charge-shielding effect of Zn2+ [26,27,28,29,30,31,32,33,34]. Therefore, OIHMs are considered as the alternatives to cathodes, which evokes a new topic of whether these OIHMs can be used to fabricate air-rechargeable ZIBs. In particular, the most of reported OIHMs with higher reduction potential (> 0.9 V vs. Zn/Zn2+) such as V2O5/PANI [26, 27], C2H8N2V7O16 [28], MnO2/PANI [29], V2O5@PEDOT [30], PEDOT-NH4V3O8 [31], NVO-PEDOT [32], VO2/PPy [33] and VOPO4/PA [34] are not conducive to fast and reliable air recharging by oxygen. Therefore, up to now, the air-rechargeable performance of OIHMs with the presence of oxygen has never been reported.

Herein, we propose an air-rechargeable ZIB based on polyaniline-coated MoS2 cathodes (MoS2/PANI). The introduction of conductive polymer (PANI) coating of MoS2 not only improves the conductivity, but also induces the charge redistribution and structure changes at the interface to weaken the electrostatic interaction, thus promoting the diffusion of Zn2+. Accordingly, MoS2/PANI cathodes show the best electrochemical performance of zinc ion storage (304.98 mAh g−1 in N2 and 351.25 mAh g−1 in air at 0.50 A g−1) among the MoS2-based cathodes so far [25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42]. More importantly, the ex situ characterization of MoS2/PANI and MoS2 proved that PANI coated on MoS2 surface could play a desolvation role, which enables MoS2/PANI materials to have great capacity and minimal lattice strain in the discharging process. Benefited from the low reduction potential of the MoS2/PANI cathodes (0.71 V vs. Zn/Zn2+), the ZIB can be quickly and deeply air recharged by oxygen (316.09 mAh g−1 at 0.50 A g−1 after 24-h air recharging, i.e., 89.99% capacity retain of galvanostatic charge at 0.50 A g−1) and have long air recharging/galvanostatic discharging life (50 cycles). As a proof of concept, the quasi-solid zinc ion batteries (QSZIBs) assembled by MoS2/PANI cathode, zinc nanoflakes anode and polyacrylamide/polyethylene glycol/zinc trifluoromethanesulfonic acid (PAM/PEG/Zn(CF3SO3)2) hydrogel electrolyte demonstrates great electrochemical performance, excellent flexibility, high- and low-temperature stability and air-recharging ability. Finally, to further verify their practicality, a 3 × 3 battery module was successfully assembled and powered a pressure sensor and a smartphone. This work will provide a promising research direction for the material design and device assembly of the next-generation self-powered system.

2 Results and Discussion

2.1 Design Principle and Structural Characterizations

The structure design, air recharging and galvanostatic discharging processes of the QSZIBs are shown in Fig. 1a. The QSZIBs consist of MoS2/PANI cathode, zinc nanoflakes anode and PAM/PEG/Zn(CF3SO3)2 hydrogel electrolyte. When discharged QSZIBs are exposed to air, Zn2+ was extracted from cathode in the presence of O2 and H2O. When QSZIBs is used to power the external circuit, the solvated Zn2+ moves from anode to cathode and is embedded into the cathode after being de-solvated on the cathode surface. The original MoS2 and MoS2/PANI nanoflowers were synthesized by simple hydrothermal method, and the successful synthesis was proved by a series of characterization. Figure 1b exhibits the X-ray diffraction (XRD) patterns of the original MoS2 (JCPDS No. 37–1492) and MoS2/PANI [43, 44]. It can be seen that compared with the original MoS2, the peak of MoS2/PANI shows a small angle shift, indicating that the most of PANI is coated on the surface of MoS2 and only a very extreme amount of PANI enters in interlayer. However, the diffraction peaks of PANI are not observed in the samples (Fig. S1a), probably because only a very small amount of PANI was introduced in the MoS2/PANI. The TG curves of MoS2 and MoS2/PANI in Fig. S2 indicate that the content of PANI in the MoS2/PANI is 5.27 wt%. The Raman spectra of original MoS2 and MoS2/PANI are shown in Fig. 1c. The interval between the E2g1 and A1g peaks of MoS2/PANI decreased, indicating that the introduction of PANI weakened interaction between neighboring MoS2 layers [35]. The FTIR and XPS results further confirmed the successful synthesis of MoS2/PANI. As shown in Figs. 1d and S1b, the FTIR spectrum of MoS2/PANI nanoflowers also shows the peaks related to PANI, but with some offsets due to the presence of MoS2. As shown in Fig. S3, the full X-ray photoelectron spectrometer (XPS) spectrum of MoS2/PANI shows the signal of Mo, S, O, C and N which are attributed to MoS2 and PANI, respectively. As can be seen from the Mo 3d XPS spectra in Fig. S4a, the peak values of 228.9 and 232.1 eV with the spin energy separation of 3.2 eV were observed, which correspond to the Mo 3d5/2 and Mo 3d3/2 of Mo4+, respectively [45], while the obvious peak at 236.2 eV is associated with the Mo6+ due to MoS2 defects caused by the introduction of PANI. It is worth noting that there is also a weak Mo6+ peak in the original MoS2 (Fig. S5a), which may be related to the incomplete decomposition of raw materials [36]. The XPS spectrums spectra were fitted to analyze the phase composition of the original MoS2 and MoS2/PANI, which are both composed of 1 T-phase and 2H-phase (Figs. S4b, S5b and Table S1). It is worth noting that the strain of lattice mismatch caused by a small amount of PANI entering the interlayer of MoS2 is the reason that MoS2/PANI contains more 1 T phases than the original MoS2. The morphology and microstructure of original MoS2, PANI and MoS2/PANI were characterized by scanning electron microscopy (SEM). Compared with the original MoS2 and PANI (Fig. S6a, b), the MoS2/PANI exhibits smaller and fluffier nanoflowers (Fig. S6c). The microstructure of original MoS2, PANI and MoS2/PANI was further studied by transmission electron microscopy (TEM). The PANI consisting of C and N shows an amorphous phase (Fig. S7). As can be seen from the TEM images of MoS2/PANI nanoflowers in different magnifications in Figs. S8a and 1e, the MoS2/PANI nanoflowers are assembled from nanosheets around 100 nm, which display much thinner and smaller nanoflowers than the original MoS2 (Fig. S9a, b). As shown in Figs. 1f and S9c, the high-resolution TEM (HRTEM) images show that the layer spacing of MoS2 (002) in MoS2/PANI increased from 0.620 to 0.682 nm, indicating that an extreme small amount of PANI entered in MoS2 layers, which is similar to the XRD results (Fig. 1b). Meanwhile, the HAADF-TEM images of original MoS2 and MoS2/PANI visually prove that the phase species are both composed of 1 T-phase and 2H-phase (Figs. S8d and S9d). Typical interfaces of MoS2 and PANI are also observed in Fig. S8b-d, which provides compelling evidence for the coating of this PANI. The STEM elemental mapping image in Fig. 1g reveals the homogeneous distribution of Mo, S, C, N and O elements along the MoS2/PANI nanoflowers.

Fig. 1
figure 1

Structure of air-rechargeable QSZIBs and characterization of MoS2/PANI nanoflowers. a Structure of air-rechargeable QSZIBs, b XRD patterns, c Raman patterns and d FTIR spectrum of original MoS2 and MoS2/PANI. e Enlarged TEM image, f HTEM image and g STEM elemental mapping image of MoS2/PANI nanoflowers

Full size image

2.2 Electrochemical Performances and Electrode Process Kinetic

The electrochemical performance of MoS2/PANI was evaluated by a standard two-electrode system in air. The CV curves at 1.0 mV s−1 and GCD curves at 2.0 A g−1 of original MoS2, PANI and MoS2/PANI are shown in Figs. 2a, b and S10a, b. The MoS2/PANI cathodes show the maximum CV area and specific capacity, indicating the excellent modification effect of PANI on MoS2. The Nyquist plot, equivalent circuit and corresponding resistances of the original MoS2, PANI and MoS2/PANI are shown in Fig. S11a, b and Table S2, respectively, indicating that the introduction of PANI can effectively improve the electrochemical reaction kinetics of MoS2. As shown in Fig. S12, we further studied the MoS2/PANI cathodes with different amounts of PANI and found that both too high and too low amount of PANI lead to the deterioration of MoS2/PANI performance, which may be due to the fact that more or less PANI is not conducive to the formation of MoS2/PANI nanoflowers (Figs. S13 and S14). The specific capacity of the above samples was further tested at current densities from 0.50 to 10.00 A g−1 (Figs. S15-S18 and 2c) to prove that MoS2/PANI cathodes had the best electrochemical performance. In addition, Fig. 2c also shows the excellent rate capability of MoS2/PANI cathode (the specific capacity at 0.50, 1.00, 2.00, 3.00, 4.00, 5.00 and 10.00 A g−1 is 351.25, 287.56, 239.93, 205.80, 190.84, 167.78 and 104.44 mAh g−1, respectively). As notably highlighted in Fig. 2d, the MoS2/PANI cathodes manifest higher energy densities over a wider power density range compared to the recently reported MoS2-based cathodes for AZIBs, which has the power density and energy density as high as 6500.0 W kg−1 and 228.3 Wh kg−1, respectively. To illustrate their excellent stability, the MoS2/PANI cathodes were tested at a current density of 5.00 A g−1. After 1500 charge/discharge cycles, the capacity retention rate of the MoS2/PANI cathodes was 79.02% (Fig. 2e). The morphology and structure of MoS2/PANI cathodes before and after the cycle were characterized by SEM to prove the stability of MoS2/PANI cathodes. Figure S19b shows SEM images of MoS2/PANI cathodes after 1500 GCD cycles (charged to 1.40 V), which are basically unchanged from original morphology and structure (Fig. S19a), indicating that the ion insertion/extraction behavior during the cycle is reversible.

Fig. 2
figure 2

Electrochemical performances and electrode process kinetic of MoS2/PANI cathodes in air. a CV curves at a scan rate of 1.0 mV s−1 and b GCD curves at a current density of 2.0 A g−1 of the MoS2/PANI cathodes and MoS2 cathodes. c Rate capability, d energy and power density plot, e cyclic life and Coulomb efficiency and f CV curves at different scan rates from 0.50 to 2.50 mV s−1 of the MoS2/PANI cathodes. g Log (i) versus log (v) plots and h contribution ratios of diffusion-controlled and capacitive capacities at different scan rates of the MoS2/PANI cathodes

Full size image

Then the energy storage mechanism of MoS2 and MoS2/PANI was studied. As shown in Figs. 2f and S20a, the electrochemical reaction kinetics of MoS2 and MoS2/PANI cathodes were revealed by CV curves at different sweep rates from 0.50 to 2.50 mV s−1. The electrochemical reaction kinetics can be assessed by Eq. 1 [46, 47]:

$$ i \, = \, av^{b} $$
(1)

As shown in Figs. 2g and S20b, the b values reveal that the electrochemical reaction kinetics of both MoS2 and MoS2/PANI cathodes are composed of capacitive and diffusion controlled. At a specified scan rate, the capacitive contribution and the diffusion-controlled contribution can be further evaluated according to Eq. 2 [48, 49]:

$$ i\left( v \right) \, = \, k_{1} v \, + \, k_{2} v^{1/2} $$
(2)

As shown in Figs. 2h and S20c, with the increase in scan rate, the ratio of the capacitive contribution process increases gradually, and the capacitive process plays a leading role in the total process. It is worth noting that the diffusion-controlled contribution of MoS2/PANI is always greater than that of MoS2 at all scan rates, indicating that the high capacity of MoS2/PANI is mainly due to ion diffusion rather than ion adsorption (Fig. S21).

2.3 Galvanostatic Intermittent Titration Technique and Density Functional Theory

The reaction kinetics of MoS2/PANI cathodes can be further studied by GITT (Fig. 3a) to estimate the Zn2+ diffusion coefficient (DZn). As shown in Fig. 3b, the DZn calculated by galvanostatic intermittent titration technique (GITT, calculation details are described in Experimental Section) in the charge and discharge process consistently reveals a DZn values range of 10−11–10−9 cm2 s−1. The DZn values achieved in this work are higher than MoS2 and previously reported manganese and vanadium-based oxide cathodes for AZIBs (Fig. S22 and Table S3). The reaction kinetics of MoS2 and MoS2/PANI were also studied by density functional theory (DFT) calculation. The disadvantage of MoS2 as cathode material of ZIBs is its weak reaction kinetics due to its poor conductivity. The initial model establishment of MoS2 and corresponding density map are shown in Fig. 3c, d, respectively. It can be seen that the original MoS2 has a relatively large band gap, which is similar to the previous report [50,51,52]. Therefore, the transfer of electrons from valence to conduction band requiring a lot of energy indicates its poor conductivity, which limits its performance. The model establishment of MoS2/PANI and corresponding density map are shown in Fig. 3e, f, respectively. It can be seen that the state density near the Fermi level of MoS2/PANI increases, electrons are easier to transition from the valence band to the conduction band, and the introduction of PANI enhances the conductivity of MoS2/PANI. As shown in Fig. 3g, h of differential charge diagram of MoS2/PANI, the PANI and MoS2 base have obvious charge transfer, which further indicates that PANI can have relatively strong interaction with MoS2 after insertion. Therefore, after the incorporation of PANI, PANI will combine with the substrate material MoS2 and further improve the conductivity of the material. In addition, DFT calculation also proves that the increase of the layer spacing reduces the zinc ion transport barrier from 0.99 to 0.39 eV (Fig. S23).

Fig. 3
figure 3

GITT and DFT of MoS2/PANI cathodes. a GITT and b Dzn of MoS2/PANI cathodes. c The initial model establishment of MoS2 and d corresponding density map. e The model establishment of MoS2/PANI and f corresponding density map. g, h The differential charge diagram of MoS2/PANI

Full size image

2.4 Energy Storage Mechanism

In order to clarify the energy storage mechanism, MoS2/ PANI cathode was tested in N2 to exclude external interference factors (such as air recharging). As shown in Figs. S24 and S25, the MoS2/PANI cathodes exhibit better cycling stability and lower capacity in N2. However, for layered materials, the interlayer spacing determines the storage capacity of zinc ions (Fig. S26), so MoS2/PANI cathodes may have some special energy storage mechanisms. In order to clarify the energy storage mechanism of MoS2/PANI cathode, the structural evolution of the MoS2/PANI cathode was characterized by ex situ XRD. The marked states (points A–E) during charge/discharge were selected for ex situ XRD tests (Fig. 4a). When the MoS2/PANI cathode discharged, the peak of (002) gradually shifts toward a smaller angle (Fig. 4b), reflecting the interlayer expansion caused by Zn2+. During charging (from state A to state C), the peak of (002) gradually recovers to the initial position, indicating that the ions insertion/extraction of MoS2/PANI cathode was highly reversible. It is worth noting that during the discharge process the peak of (002) moves very a little toward the lower angle, indicating that the interlayer expansion of MoS2/PANI cathode during charging and discharging process is very small. Generally speaking, when materials are discharged, they exhibit small interlayer expansion mainly for the following reasons: (1) limited zinc ion embedding; (2) other small ions interlayer contribution capacity; (3) there is desolvation phenomenon. First, the Inductive Coupled Plasma Emission Spectrometer (ICP) analyses of MoS2/PANI cathode at fully discharged and charged were tested. As shown in Table S4, the ratio of Zn to Mo atoms before and after MoS2/PANI charging and discharging is 0.8474 and 0.0237, which proves that the capacity of MoS2/PANI is mainly contributed by zinc ion intercalation. Then, we verify that the H+ insertion has a little contribution to the cathode capacity (Fig. S27). However, the presence of Zn2+ in 2.0 M Zn(CF3SO3)2 takes the form of larger Zn(CF3SO3)2(H2O)4 (inserted form is Zn(H2O)42+), which usually causes larger interlayer expansion when inserted in electrode materials with smaller interlayer spacing [53,54,55,56]. Therefore, we supposed that when Zn(H2O)42+ is inserted in MoS2/PANI cathode, there is a mechanism of desolvation on the surface of electrode material, which causes Zn(CF3SO3)2(H2O)4 to change into smaller Zn2+. Therefore, further tests were carried out to prove above conjecture and clarify the storage mechanism of MoS2/PANI cathodes. The ex situ Raman spectra of MoS2/PANI cathode at different charge and discharge states were collected (Fig. 4c). When MoS2/PANI cathode was discharged to 0.10 V, the intensity ratio of A1g/E12g decreased (A and E), indicating that ions had been inserted into the interlayer of MoS2/PANI cathode. When MoS2/PANI cathode was charged to 1.40 V, the intensity ratio of A1g/E12g recovered to 1.47 due to the extraction of ions. The intensity ratio of A1g/E12g changes slightly during charging and discharging process of MoS2/PANI cathode, which are consistent with the results of ex situ XRD. In order to prove that Zn2+ rather than other ions inserted MoS2/PANI cathode, XPS spectrum of the original, fully discharged and charged states has been studied. As shown in Fig. 4d, no Zn signal was detected in the Zn 2p XPS spectrum of the original electrode. When MoS2/PANI cathode was fully discharged, two pairs of Zn 2p signals can be observed, corresponding to the Zn2+ absorbed on the electrode surface and the Zn2+ inserted in electrode, respectively. After fully charging to 1.40 V, a pair of weak Zn signals were still observed due to the adsorption of residual Zn(CF3SO3)2. Spectra and fitting results of XPS in different states in Mo 3d of MoS2/PANI cathode are shown in Fig. 4e and Table S1. The Mo 3d XPS deconvolution shows that the pristine electrode contains 65.82% 1 T-MoS2 and 29.18% 2H-MoS2, along with 5.00% Mo6+ caused by surface oxidation. After being discharged to 0.10 V, the proportion of 2H-MoS2 is reduced to 18.78% (against 65.74% for 1 T-MoS2 and 15.48% for Mo6+), illustrating the phase transition from 2H- to 1 T-MoS2 induced by Zn2+ insertion. It is worth noticing that the 1 T-MoS2 could be easily oxidized (during the XPS testing process), especially after electrochemical activation [29], which explains the dramatically strengthened Mo6+ signal and the reduced 1 T-phase content. After charging to 1.40 V, the original content of 2H-MoS2 is finely restored, indicating the highly reversible phase transition triggered by Zn2+ insertion/extraction. The reversible phase transition between 1 T-MoS2 and 2H-MoS2 during Zn2+ insertion/extraction was confirmed. The ex situ HRTEM image (Fig. 4f, g) visually shows the fringe distance of 0.719/0.684 nm at the insertion/extraction of Zn2+, which is in good agreement with the XRD results. STEM elemental mapping images reveal the uniform distribution of Mo, S, C, N, O and Zn elements in the fully discharged electrode (Fig. 4h) and fully charged electrode (Fig. 4i), indicating the reversible insertion/extraction of Zn2+. Based on the above discussion, the storage mechanism of MoS2/PANI cathode is summarized as the reversible insertion/extraction of Zn2+ formed by the desolvation of Zn(CF3SO3)2(H2O)4 on the electrode surface rather than Zn(H2O)42+. To investigate whether PANI plays an important role in the desolvation of Zn(CF3SO3)2(H2O)4, the Zn storage mechanism of the original MoS2 electrode has also been studied (Figs. S28 and S29). It is found that the peak of (002) has obvious shift to lower angle and the layer spacing increases to 0.962 nm. In addition, it is noteworthy that no significant change in the peak intensity of F 1 s XPS spectrum was detected in the full charging and discharging of original MoS2 and MoS2/PANI cathode, indicating that CF3SO32− really did not inserted in the electrode and the detected F 1 s spectra were attributed to the adsorption of electrolyte (Fig. S30). However, when the original MoS2 cathode was fully charged and discharged, the detected O 1 s XPS spectral peak intensity changed significantly, indicating that Zn(H2O)42+ may be inserted into the electrode (Fig. S31). Energy storage mechanism of original MoS2 and MoS2/PANI cathode in discharge process is shown in Fig. 4j, k. For the original MoS2, Zn(H2O)42+ formed by the decomposition of Zn(CF3SO3)2(H2O)4 on the surface of MoS2 was directly inserted into MoS2, resulting the layer spacing increases. For MoS2/PANI cathode, Zn(CF3SO3)2(H2O)4 will first form coordination with PANI on the surface. Benefits from the synergistic effect of Zn2+ and π electrostatic interaction, the solvent shell around the Zn2+ will collapse in the transmission process [57, 58]. Finally, Zn2+ are separately extracted into the cathode material through PANI (Fig. S32). Therefore, MoS2/PANI cathode has a very small layer spacing increases after discharge and provides a very large capacity. Based on the above results and discussion, the overall electrochemical reaction of the ZIBs based on MoS2/PANI cathode can be described as follows:

Fig. 4
figure 4

Zn storage mechanism of MoS2/PANI cathodes. a Initial discharge/charge curve at 0.50 A g−1, the marked states are selected for ex situ tests. Ex situ b XRD patterns, c Raman patterns, XPS spectra of d Zn and e Mo. f, g HRTEM and h, i STEM element mapping images of the fully discharged and charged MoS2/PANI cathodes. j, k The solvation and desolvation of MoS2 and MoS2/PANI

Full size image

Cathode:

$$ {\text{MoS}}_{{2}} /{\text{PANI }} + \, 0.{\text{85 Zn}}^{{{2} + }} + { 1}.{\text{7e}}^{ - } \leftrightarrow {\text{ MoZn}}_{{0.{85}}} {\text{S}}_{{2}} /{\text{PANI}} $$
(3)

Anode:

$$ 0.{\text{85 Zn}} \leftrightarrow 0.{\text{85Zn}}^{{{2} + }} + {1}.{\text{7e}}^{ - } $$
(4)

Overall:

$$ {\text{MoS}}_{{2}} /{\text{PANI }} + 0.{\text{85 Zn }} \leftrightarrow {\text{ MoZn}}_{{0.{85}}} {\text{S}}_{{2}} /{\text{PANI}} $$
(5)

2.5 Air-Rechargeable Mechanism

The MoS2/PANI cathode exhibits excellent Zn2+ storage performance, and the Mo was reduced/oxidized during the insertion/extraction of Zn2+. During electrochemical charging, electrons are released from the MoZn0.85S2/PANI (Zn0.85Mo/P) cathode and Mo in Zn0.85Mo/P was oxidized and Zn2+ was extracted from the layered structure. In this process, the driving force for Zn0.85Mo/P to release electrons is generally an external power supply. In addition to the above electrochemical oxidation reactions, other strategies that can realize electron transfer are also expected to be used in the charging process of air-rechargeable ZIBs based on Zn0.85Mo/P cathode. Among various oxidants, O2 is common and abundant in air and has the standard electrode potentials of ~ 0.40 and ~ 1.23 V vs. SHE in the neutral and acidic medium, respectively. Therefore, if the redox potential of Zn0.85Mo/P cathode is lower than O2, the oxidation of Mo and the extraction of Zn2+ of Zn0.85Mo/P cathode can be realized. In order to verify the spontaneity of the redox reaction between Zn0.85Mo/P and O2 in 2.0 M Zn(CF3SO3)2 electrolyte, a galvanic cell was designed. As shown in Fig. 5a, the Zn0.85Mo/P electrode served as anode with 2.0 M Zn(CF3SO3)2 electrolyte and was sealed under N2 to avoid the dissolvation of oxygen. The platinum sheet was used as cathode, which was immersed in 2.0 M Zn(CF3SO3)2 electrolyte containing dissolved oxygen. In this system, a cell voltage of 0.499 V was observed. According to the relationship between thermodynamic function and cell voltage:

$$ \Delta G = - nEF $$
(6)

where ΔG, F, E and n are the Gibbs free energy change, Faraday constant cell voltage and normal number, respectively, the Gibbs free energy change of above system is less than zero, indicating that the redox reaction between Zn0.85Mo/P and O2 can take place spontaneously in 2.0 M Zn(CF3SO3)2.The air-recharging mechanism of ZIBs based on Zn0.85Mo/P cathode is shown in Fig. 5b. Under the action of O2, Zn2+ is extracted from Zn0.85Mo/P, while O2 gains electrons from Zn0.85Mo/P and reacts with H2O to form OH−. The generated OH− and extracted Zn2+ combine with the adsorbed electrolyte ions (Zn2+ and CF3SO32−) to form an amorphous trifluoride containing Znx+y(CF3SO3)2y(OH)2x [Figs. S33-S35] [18]. Therefore, the fully discharged Zn0.85Mo/P cathode can be restored to the charged state Zn0.85-xMo/P cathode through spontaneous redox reaction. Subsequently, the curve of open-circuit voltage (OCV) with and without O2 as a function of air-recharging time was tested (Fig. 5c), which further prove that O2 plays an important role. More importantly, it is found that after 24.0 h of oxidation, the OCV of the ZIBs reached 1.15 V, close to the initial OCV (1.21 V). In order to demonstrate the above mechanisms, ex situ characterization was carried out to study the structure and composition evolution of Zn0.85-xMo/P electrodes. As the air-rechargeable time is extended from 0 to 24.0 h, the XRD peak (002) of Zn0.85-xMo/P electrodes gradually shifts to a large angle (Fig. 5d), and the intensity ratio of A1g/E12g in Raman spectrum increases from 1.05 to 1.39 (Fig. 5e). This result is similar to that of chemical charging, indicating that Zn2+ gradually extracted from Zn0.85-xMo/P electrode. In order to further prove the air-recharging mechanism of Zn0.85-xMo/P electrode, XPS was carried out. With the increases of air-recharging time, the content of Zn in Zn0.85-xMo/P electrode gradually decreased (Fig. 5f) and the Mo 3d peak gradually moved to a higher binding energy (Fig. 5g), indicating the gradual extraction of Zn2+ and oxidation of Mo in Zn0.85-xMo/P cathode during the air-recharging process. The above discussion proves that the air-rechargeable ZIB based on Zn0.85-xMo/P cathode can be charged by O2. In order to determine the air-recharging performance of ZIB based on Zn0.85-xMo/P cathode, the galvanostatic current discharge curves after air recharging for different times were tested (Fig. 5h). Since the air-rechargeable battery needs to consume zinc electrodes continuously during its operation, the mass of the zinc anode is strictly controlled at 21.2 mg to assemble the battery for practicability. After air recharging for 4.0 h, the corresponding discharge capacity of the ZIB can reach to 64.17 mAh g−1 at 0.50 A g−1. As the air-recharging time increases, the capacity of the ZIB gradually increases. 35 Rechargeable ZIB could provide 316.09, 212.69, 164.94, 146.50, 133.61, 123.47 and 85.83 mAh g−1 at the current densities of 0.50, 1.00, 2.00, 3.00, 4.00, 5.00 and 10.00 A g−1, respectively. As shown in Fig. S36, the air-rechargeable ZIB manifests high energy densities over a wider power density range which has the power density and energy density as high as 288.29 W kg−1 and 12.69 Wh kg−1, respectively. The stability of this air recharging/galvanostatic current discharge cycle was subsequently verified. In this test, the air-recharging ZIB based on Zn0.85-xMo/P cathode was discharged to 0.10 V at a current density of 0.50 A g−1 and then air recharged for 24.0 h and galvanostatic discharge at 0.50 A g−1. As shown in Fig. 5j, the air-rechargeable ZIB can be cycled for 50 times. Afterward, the air-rechargeable ZIB shows a high discharge capacity of 291.22 mAh g−1 with the high capacity retention rate of 82.91%. Based on the above discussion, the redox reaction between Zn0.85Mo/P and O2 can be summarized as below:

$$ \begin{aligned}&{\text{MoZn}}_{{0.{85}}} {\text{S}}_{{2}} /{\text{PANI }} + \, 0.{5}x{\text{O}}_{{2}} + \, x{\text{H}}_{{2}} {\text{O }} + \, y{\text{Zn}}\left( {{\text{CF}}_{{3}} {\text{SO}}_{{3}} } \right)_{{2}} \to {\text{MoZn}}_{{0.{85} - x}} {\text{S}}_{{2}} /{\text{PANI }} \\& + {\text{ Zn}}_{x + y} \left( {{\text{CF}}_{{3}} {\text{SO}}_{{3}} } \right)_{{{2}y}} \left( {{\text{OH}}} \right)_{{{2}x}} \end{aligned}$$
(7)
Fig. 5
figure 5

Mechanism of the redox reaction between Zn0.85Mo/P and O2 and air recharging/galvanostatic discharging behavior of Zn//Zn0.85−xMo/P batteries. a Optical image of the designed galvanic cell. b Working mechanism of air-rechargeable ZIBs. c Effect of the oxidation time on OCV of Zn//Zn0.85−xMo/P batteries. d XRD patterns, e Raman patterns, XPS spectra of f Zn and g Mo of Zn//Zn0.85−xMo/P batteries after different air-recharging time. h Galvanostatic discharging curves of Zn//Zn0.85−xMo/P batteries after different air-recharging time. i Galvanostatic discharging curves of Zn//Zn0.85−xMo/P batteries at different current density after 24.0-h air recharging. j Air recharging/galvanostatic discharging life of Zn//Zn0.85−xMoS batteries

Full size image

2.6 Practicability and Zn Batteries Module

In order to illustrate the practicability of MoS2/PANI cathode, QSZIBs were assembled by MoS2/PANI cathode, zinc nanoflakes synthesized on carbon cloth by electrodeposition (Fig. S37) to assemble the battery for practicability. Anode and PAM/PEG/Zn(CF3SO3)2 served as hydrogel electrolyte (Fig. 6a). As shown in Fig. 6b, the QSZIB shows excellent electrochemical performance. The discharge capacities at 0.50, 1.00, 2.00, 3.00, 4.00 and 5.00 A g−1 are 300.96, 248.04, 162.47, 123.63, 93.44 and 68.06 mAh g−1, respectively. Considering that batteries often need to be operated in extreme conditions, QSZIB was tests under bent or high and low temperatures (Fig. 6c). The QSZIB shows only slight capacity variation under different bending states, indicating the excellent flexibility of the battery. Even more impressively, the QSZIB demonstrated high capacity and stability at both high and low temperature, respectively. Subsequently, the air-recharging ability of QSZIB was verified. After air recharging for 24.0 h, the fully discharged QSZIB shows a high capacity at different current densities (Fig. 6d). Considering the low capacity and voltage of QSZIB caused by the inherent characteristics of water batteries, a 3 × 3 module was constructed to further verify the practicality (Fig. 6e). In subsequent tests, the module was found to successfully power pressure sensors (Fig. 6f) and smartphones (Fig. 6g), showing great practical prospects.

Fig. 6
figure 6

Practicability of MoS2/PANI cathode. a Schematic diagram and b GCD curves of the QSZIBs. c Specific capacity at 2.00 A g−1 of QSZIBs under bent or high and low temperatures. d Galvanostatic discharging curves of QSZIBs after 24.0-h air recharging. e GCD curves at 2.0 A g−1 of QSZIB module. The QSZIB module powered for f pressure sensors and g smartphones

Full size image

3 Conclusions

In summary, we design a air-rechargeable ZIB based on MoS2/PANI cathode. The introduction of conductive polymer (PANI) coating of MoS2 not only improves the conductivity, but also induces the charge redistribution and structure changes at the interface to weaken the electrostatic interaction, thus promoting the diffusion of Zn2+. Accordingly, MoS2/PANI cathodes show the best electrochemical performance of zinc ion storage among the Mo-based cathodes so far. Benefited from the low reduction potential of the MoS2/PANI cathodes (0.71 V vs. Zn/Zn2+), the ZIBs can be quickly and deeply air recharged by oxygen (316.09 mAh g−1 at 0.50 A g−1 after 24-h air recharge, i.e., 89.99% capacity retain of galvanostatic charge at 0.50 A g−1) and have long air recharging/galvanostatic discharging life (50 cycles). As a proof of concept, the quasi-solid-state ZIBs employing MoS2/PANI cathode and PAM/PEG/Zn(CF3SO3)2 demonstrate great electrochemical performance, excellent flexibility, high- and low-temperature stability and air-rechargeable ability. Finally, to verify the practicality of the battery, a 3 × 3 battery module was successfully assembled to power a pressure sensor and a smartphone. This work will provide a promising research direction for the material design and device testing of the next-generation self-powered system.