Introduction

Wearable electronics have attracted considerable interest in the fields of health care and human-machine interfaces because of their flexibility and portability [1]. Due to the direct contact between wearable electronics and the body, the energy storage devices, such as batteries, that supply the power are required to be miniaturized, flexible, and mechanically matched (e.g., the Young’s moduli) with biological tissues to avoid physical irritation and internal injury [2, 3]. Conventional and bulky three-dimensional batteries and two-dimensional planar structures are typically rigid and cannot meet the above requirements, and this has become a bottleneck for further development of wearable electronics [4, 5]. In contrast, one-dimensional fiber batteries are smaller and adapt effectively to complex deformations, making them suitable candidates for powering wearable electronics [6, 7].

However, available fiber batteries do not possess mechanical properties (e.g., Young’s moduli) that match those of biological tissues since the materials employed are intrinsically rigid or have limited flexibility. For example, metal-air fiber batteries with anodes containing metallic wires made of lithium (Li), magnesium (Mg), or zinc (Zn) exhibited Young’s moduli of 107–108 kPa [8,9,10]. Metal-ion fiber batteries are endowed with certain flexibility owing to the use of carbon materials such as aligned carbon nanotube (CNT) sheets and reduced graphene oxide fibers as current collectors. However, the Young’s moduli of these batteries are still too high (106–108 kPa) [11, 12]. Although elastic electrodes designed recently from elastomers have decreased the Young’s moduli of batteries to 104–105 kPa [13], their rigidity is still orders of magnitude higher than that of biological tissues, which have Young’s moduli of 101–103 kPa [14, 15]. To meet the demands of compatible mechanical properties completely, it is essential to fabricate fiber batteries entirely from ultrasoft materials while not sacrificing their softness during integration. Currently, ultrasoft materials such as hydrogels are used as electrolytes to fabricate flexible batteries, but the electrodes or packing materials are rigid and reduce the overall softness [4, 8,9,10, 12]. In our previous work, we successfully utilized an all-hydrogel design to develop planar lithium-ion batteries [16]. However, use of an all-hydrogel design to fabricate fiber batteries has not been reported and remains challenging. Here, to the best of our knowledge, the first ultrasoft aqueous Li-ion batteries with coaxial fiber structures have been fabricated with an all-hydrogel design. Intimate and stable interfaces formed between the hydrogel electrodes and hydrogel electrolyte ensured that the batteries exhibited high electrochemical performance. The integrated batteries exhibited a high specific discharge capacity of 84.8 mAh·g−1 and stable cycling behavior and rate capacity at a current density of 0.5 A·g−1. Moreover, the all-hydrogel design endowed the batteries with considerable softness. The batteries showed a low Young’s modulus of 445 kPa and perfectly matched the mechanical properties of biological tissues. Stable interfaces were achieved between the electrodes and electrolyte with stable interfacial adhesion and impedance, and the electrochemical properties were maintained during and after different complex deformations.

Experimental procedure

Materials

Acrylamide (AAM, 99.0%), lithium chloride (LiCl, 99.0%), and methylene-bis-acrylamide (MBA, 99.0%) were purchased from Shanghai Aladdin Biochemical Technology Co. Ltd. Ammonium persulfate (APS, 98.0%) and was purchased from Nanjing Chemical Reagent Co. Ltd. N, N, N’, N’-tetramethylethylenediamine (TEMED, 99.0%, Alfa Aesar) was purchased from Shanghai Sarn Chemical Technology Co. Ltd. LiMn2O4 (LMO) was purchased from Shanghai Aidu Chemical Technology Co. Ltd., and LiTi2(PO4)3 (LTP) was purchased from Jiangxi KingLi Technology Co. Ltd. A multiwall carbon nanotube (MWCNT) aqueous dispersion (TNWDMC-MC8, 10.0 wt%) was purchased from Chengdu Organic Chemicals Co. Ltd. Polyvinylidene fluoride (PVDF), active carbon (AC), and Super P were purchased from Hefei Kejing Material Technology Co. Ltd. N-Methyl-pyrrolidone (NMP, 99.0%) was purchased from Shanghai Macklin Biochemical Co. Ltd. Heat-shrinkable tubes were purchased from Shenzhen Woer Heat-shrinkable Material Co. Ltd. Krazy glue was purchased from a supermarket. All chemicals were used as received.

Preparation of polyacrylamide (PAM)/CNT hydrogel fiber and PAM/CNT hydrogel film

First, the precursor solution for the PAM/CNT hydrogel fiber was mixed from 7.5 mL of a MWCNT aqueous dispersion, 12.5 mL of deionized water, 1.4 g of AAM, 4.0 mg of MBA, and 40.0 mg of APS. The precursor solution was stirred for 60 min to obtain a homogeneous and clear solution. After adding 40.0 μL of TEMED as a catalyst, the obtained precursor solution was immediately injected into a heat-shrinkable tube with an inner diameter of 2.8 mm by using a syringe. After free-radical polymerization for 5 min, the PAM/CNT hydrogel fiber was extracted from the tube with a syringe for further use. Likewise, the precursor solution for the PAM/CNT hydrogel film was mixed from 1.9 mL of a MWCNT aqueous dispersion, 3.1 mL of deionized water, 355.2 mg of AAM, 1.0 mg of MBA, and 10.0 mg of APS. After the precursor solution was stirred for 60 min, 10.0 μL of TEMED was added as a catalyst. The obtained solution was immediately transferred into a petri dish with an inner diameter of 55.0 mm to enable free-radical polymerization. The obtained PAM/CNT hydrogel film was saved for further use.

Preparation of the hydrogel fiber cathode and hydrogel anode

The hydrogel fiber cathode and anode were obtained by loading LMO and LTP active material dispersions onto the PAM/CNT hydrogel fiber and PAM/CNT hydrogel film, respectively. Specifically, active material (LMO or LTP), Super P, and PVDF with mass ratios of 16:3:1 were first ground in a mortar to obtain a homogeneous, active mixed powder. Then, NMP was added to the mixed powder to obtain a dispersion of active material (LMO or LTP) by further grinding and ultrasonicating for 1 min. The concentrations of LMO and LTP in their dispersions were 200.0 and 32.0 mg·mL−1, respectively. For the hydrogel fiber cathode, the LMO dispersion was dip-coated onto the PAM/CNT hydrogel fiber. The dip-coating process was repeated twice, and each lasted 10 s. For the hydrogel anode, the LTP active material dispersion was added dropwise onto the PAM/CNT hydrogel film using a pipette. The loading area of the LTP active material dispersion was 20.0 μL·cm−2. Finally, the hydrogel fiber cathode and anode were placed in a 60 °C oven for thorough dehydration and then heated in a 60 °C vacuum oven overnight to remove residual NMP.

Preparation of the hydrogel electrolyte

The precursor solution to the hydrogel electrolyte was a mixture containing 1.7 g of AAM, 1.0 g of LiCl, and 2.4 mg of MBA dissolved in 12.0 mL of deionized water mixed with magnetic stirring for 1 h. Argon was pumped into the solution for 1 h to remove oxygen. After adding 24.0 mg of APS and 8.0 μL of TEMED into the above solution and stirring for 1 min, the obtained solution was immediately transferred into a petri dish with an inner diameter of 55.0 mm to enable the free-radical polymerization reaction. The resulting PAM/LiCl hydrogel electrolyte was saved for further use.

Fabrication of an all-hydrogel fiber aqueous Li-ion battery

The all-hydrogel fiber aqueous Li-ion battery was fabricated by sandwiching the hydrogel electrolyte between the dehydrated hydrogel fiber cathode and hydrogel anode in a coaxial fiber structure. The as-fabricated all-hydrogel fiber aqueous Li-ion battery was held at room temperature for 1 h to allow the dehydrated electrodes to rehydrate to equilibrium.

Results and discussion

The coaxial-fiber configuration of the all-hydrogel fiber aqueous Li-ion battery is schematically demonstrated in Fig. 1a. The fabrication process was divided into the following steps (Fig. S1). First, the hydrogel fiber cathode and anode were made from PAM/CNT hydrogels loaded with active materials. Then, the as-prepared electrodes were fully dried in the oven into a dehydrated state. Finally, the all-hydrogel fiber aqueous Li-ion battery was fabricated by sequentially covering the PAM/LiCl hydrogel electrolyte and wrapping the dehydrated hydrogel anode onto the dehydrated hydrogel fiber cathode. We evaluated the conversions of the PAM/CNT hydrogel and PAM/LiCl hydrogel, and high degrees of monomer conversion (>90%) were achieved (Fig. S2). This confirmed that the gelation process was sufficient and beneficial in providing structural stability and the necessary mechanical properties for the hydrogels. A photograph of an as-fabricated battery with a diameter of several millimeters is shown in Fig. 1b. During the fabrication process, when the dehydrated electrodes were in contact with the water-rich electrolyte, they immediately absorbed water from the electrolyte spontaneously because of the hydrophilic amide groups on the polymer chains and osmotic pressure between the electrodes and electrolyte [17]. When the spontaneous water absorption reached equilibrium, the rehydrated electrodes were moist and became soft. In addition, amide groups on the PAM side chains of the electrodes and the electrolyte formed abundant hydrogen bonds at the interfaces (Fig. S3) [18, 19]. Fourier transform infrared spectroscopy (FTIR) results further verified formation of hydrogen bonds (Fig. S4). In addition to hydrogen bonds, van der Waals interactions between polymers were beneficial in maintaining the stability of the interface [20]. Using these interactions, intimate and stable interfaces were constructed between the rehydrated electrodes and electrolyte (Fig. 1c, d). The tightly adhering interfaces might account for the strong electrochemical performance of the batteries.

Fig. 1
figure 1

Schematic illustration and structural characterization of all-hydrogel fiber aqueous Li-ion batteries. a Illustrations of the coaxial fiber structure and the interface between the rehydrated hydrogel electrode and hydrogel electrolyte. b Photograph of the all-hydrogel fiber aqueous Li-ion battery. c SEM image of the interface between the rehydrated hydrogel fiber cathode and hydrogel electrolyte. d SEM image of the interface between the rehydrated hydrogel anode and hydrogel electrolyte

The electrochemical performance of an all-hydrogel fiber aqueous Li-ion battery was subsequently investigated. The PAM/LiCl hydrogel was used as the hydrogel electrolyte. The ionic conductivity was 38 mS·cm−1 (Fig. S5), which was comparable to those reported for hydrogel electrolytes used in aqueous Li-ion batteries (Table S1). The hydrogel fiber cathode and anode were prepared by loading the LMO and LTP active materials, respectively (Figs. S6 and S7). The active materials on the hydrogel fiber cathode and hydrogel anode were both distributed uniformly, which favored the reactions of active materials and enhanced the electrochemical performance of the batteries. The mass ratio of LMO to LTP was chosen to be 0.8:1 to ensure efficient utilization of materials. To investigate the electrochemical performance of the all-hydrogel fiber aqueous Li-ion battery, we first obtained cyclic voltammograms (CVs) for the PAM/CNT-LMO and PAM/CNT-LTP hydrogel electrodes (Figs. S8 and S9) with a three-electrode system. For the PAM/CNT-LMO hydrogel fiber cathode, two pairs of redox peaks located at 0.90/0.70 V and 1.05/0.82 V (vs. Ag/AgCl) were observed for LMO. These results showed that intercalation and deintercalation of Li+ were reversible before evolution of oxygen. For the PAM/CNT-LTP hydrogel anode, the redox reactions were reversible without substantial hydrogen evolution. Here, CVs of the all-hydrogel fiber aqueous Li-ion battery were obtained by cycling between 0 and 1.80 V at a scan rate of 0.1 mV·s1 with a two-electrode system (Fig. 2a). For the battery, the characteristic redox peaks were at 1.20 V/1.17 V and 1.70 V/1.60 V, indicating that the redox reactions were reversible. To avoid electrolyte decomposition and ensure normal battery operation, the operating potential for the battery test was set to 1.80 V. As shown in Fig. 2b, the all-hydrogel fiber aqueous Li-ion battery exhibited an average discharge voltage plateau of 1.50 V, which was similar to those of conventional aqueous Li-ion batteries and remained almost unchanged over 100 cycles. The specific discharge capacity was 84.8 mAh·g−1 based on the mass of LTP at a current density of 0.5 A·g−1. The battery also exhibited high stability with 92% capacity retention over 100 cycles at a current density of 0.5 A·g−1, and nearly 100% coulombic efficiency was achieved for each cycle (Fig. 2b). As demonstrated in Fig. 2c, d, the specific discharge capacity of the battery decreased from 85.7 to 78.3, 70.6, and 63.4 mAh·g−1 with increasing current densities from 0.5 to 0.75, 1, and 1.25 A·g−1, respectively, indicating that it can be operated steadily over a wide range of current densities. Moreover, the capacity recovered to 85.7 mAh·g−1 when the current density was returned to 0.5 A·g-1.

Fig. 2
figure 2

Electrochemical performance of the all-hydrogel fiber aqueous Li-ion battery. a CV for the all-hydrogel fiber aqueous Li-ion battery cycled between 0 and 1.80 V at a scan rate of 0.1 mV·s1. b, c Galvanostatic charge‒discharge curves for the all-hydrogel fiber aqueous Li-ion battery and the corresponding cycling behavior at a current density of 0.5 A·g1. Here, C0 and C correspond to the specific discharge capacities before and after cycling, respectively. d, e Galvanostatic charge‒discharge curves for the all-hydrogel fiber aqueous Li-ion battery at the different current densities 0.50, 0.75, 1.00, and 1.25 A·g1 and the corresponding rate capacity performance

In addition to the electrochemical performance, the mechanical properties of the all-hydrogel fiber aqueous Li-ion batteries were also elucidated. Figs. 3a and S10 show typical stress‒strain curves for the battery and its components derived from standard uniaxial tensile tests. Based on calculations from the linear segments of the curves, the Young’s moduli of the hydrogel fiber cathode, hydrogel anode, hydrogel electrolyte and all-hydrogel fiber aqueous Li-ion battery were 180, 254, 12, and 445 kPa, respectively. Among the components of the batteries, the internal stress of the hydrogel fiber under compression was much smaller than those of traditional conductive metallic wires and was superior those of flexible CNT fibers (Figs. 3b and S11). This result indicated that the hydrogel fiber was soft, bendable, and likely to be highly compatible with biological tissues. Furthermore, the Young’s moduli of the all-hydrogel fiber aqueous Li-ion batteries were several orders of magnitude lower than those of various reported fiber batteries based on metallic wires, carbon materials, and elastomers (Fig. 3c). More importantly, the Young’s moduli of the batteries matched those of biological tissues, ensuring that the all-hydrogel fiber aqueous Li-ion batteries were mechanically compatible with tissues and would preclude potential physical irritation and internal injury.

Fig. 3
figure 3

Mechanical properties of the all-hydrogel fiber aqueous Li-ion batteries. a Stress‒strain curves of the rehydrated hydrogel fiber cathode, rehydrated hydrogel anode, hydrogel electrolyte and all-hydrogel fiber aqueous Li-ion battery (inset, stress‒strain curves for strains within 20%). b Internal stress of the hydrogel fiber under compression (with a length of 4 mm) in comparison with those of other conductive fibers, including Cu, Zn, Mg, Ti, stainless steel (SS) wires, and CNT fibers. c Comparison of the Young’s modulus for the all-hydrogel fiber aqueous Li-ion battery (based on PAM hydrogel) with those of metal-air fiber batteries based on metallic wires (e.g., Li, Mg, and Zn wires) [8,9,10], a metal-ion fiber battery based on carbon materials (e.g., CNT and reduced graphene oxide fiber) [11, 12], an elastic fiber battery based on elastomers (e.g., butadiene rubber (SBR)) [13], and biological tissues (e.g., skin and brain) [14, 15]

The intimate and stable interfaces formed between the rehydrated electrodes and electrolyte were also maintained perfectly during dynamic deformation. Owing to the adhesiveness of the interface and the similar mechanical properties of the components, delamination of the electrode and electrolyte during deformation was prevented. After bending to 45°, 90°, and 135° each for 1000 cycles, the adhesion strength of the interface became slightly stronger (Figs. 4a and S12), which was because the van der Waals interactions between the polymers had increased [20]. Moreover, as the number of bending cycles for each angle increased, the impedances measured under the same frequency remained almost unchanged (Fig. 4b). The charge transfer resistance of the all-hydrogel fiber aqueous Li-ion battery and those of the cathode/anode and hydrogel/electrolyte interfaces exhibited minor fluctuations (average changes within 10%) under repeated deformation, proving the stabilities of the interfaces from the perspective of charge transfer (Figs. S13 and S14). Scanning electron microscopy (SEM) images of the interfaces and X-ray diffraction data for LMO and LTP further illustrated that the material did not deteriorate and that the interfaces were stable (Figs. S15 and S16). Due to the all-hydrogel design and the stable interfaces between the electrodes and electrolyte, strong electrochemical performance and matched mechanical properties were achieved, and the batteries exhibited stable electrochemical behaviors during complex deformations, including bending and stretching. In Fig. 4c, the all-hydrogel fiber aqueous Li-ion batteries were bent to the different angles 45°, 90°, and 135°, and the capacity retention (C/C0) was approximately 100% after 100 galvanostatic charge‒discharge cycles for each bending state (Fig. S17). In determining the stability of the battery to cyclic deformation, the specific discharge capacities showed only slight declines after bending at 45°, 90°, and 135° and stretching by 30% for 1000 cycles (Fig. 4d). The stable electrochemical performance observed during complex deformation showed that the all-hydrogel fiber aqueous Li-ion batteries are suitable for powering wearable electronics.

Fig. 4
figure 4

Softness and electrochemical stability of the all-hydrogel fiber aqueous Li-ion battery. a Shear strengths of the interfaces between hydrogel electrodes and hydrogel electrolyte of all-hydrogel fiber aqueous Li-ion batteries before and after 1000 deformation cycles. b Electrochemical impedance spectroscopy after bending cycles at different bending angles of 45°, 90°, and 135°. c Cycling behavior at a current density of 0.5 A·g−1 for different bending angles. d Specific discharge capacity retention after bending at 45°, 90°, and 135° and stretching to 30% for 1000 cycles (inset, illustrations of bending at 45°, 90°, and 135° and stretching of the battery by 30%). Here, C0 and C correspond to the specific discharge capacities before and after deformation, respectively

Conclusion

To summarize, we have developed all-hydrogel aqueous Li-ion batteries with coaxial fiber architectures. The obtained batteries exhibited strong and stable electrochemical performance under complex deformations and a low Young’s modulus that matched those of biological tissues well. This work offers a promising strategy for developing advanced soft fiber batteries and provides a remedy for mechanical mismatch between fiber batteries in wearable electronics and biological tissues.