Scalable high-voltage Zn||MnO₂ batteries achieved by mild amphiphilic hydrogel electrolytes

Significance

Zn||MnO₂ batteries are over 100 y old technology and are widely used as nonrechargeable dry batteries nowadays. Rechargeable Zn||MnO₂ batteries meet the challenges of how to simultaneously achieve a high discharge plateau and large areal capacity in neutral aqueous electrolytes. In this work, we develop a mild amphiphilic hydrogel electrolyte that shows an expanded electrochemical stability window in high water content, allowing high valent conversion of Mn species. Meanwhile, the brush-like spikes of the hydrogel electrolyte effectively optimize the cathode with high mass loading. This work attempts to put forward high-voltage aqueous Zn||MnO₂ batteries to the commercialization level.

Abstract

The practical applications of aqueous Zn||MnO₂ batteries are limited by their small areal capacity, low discharging plateau, and clumsy packing device. Currently, the high potential MnO₂/Mn²⁺ redox conversion can only be well activated in electrolytes with a very low pH value, which is not friendly to the Zn metal anode. To overcome these limitations, we have designed mild amphiphilic hydrogel electrolytes (AHEs) with a wide electrochemical stability window (ESW) and high ionic activity. The design is based on the mechanism that trace amounts of hydrophobic moieties enhance the hydrogen bonding between hydrophilic groups and water molecules in the hydrogel electrolytes. The developed AHE possesses an ESW up to ~3.0 V even at a high water content of ~76 wt%. The assembled Zn||MnO₂ pouch cells using the hydrogel electrolytes demonstrated a large areal capacity of ~5 mAh cm⁻² at 1 mA cm⁻² and a high-voltage and flat discharging plateau of ~1.9 V. Furthermore, a pouch cell with an area of 40 cm² was fabricated, exhibiting a capacity of ~125 mAh at 2 mA cm⁻². Two pouch cells (25 cm²) in series were used to drive a 3.7 V-powerable electric fan. This work highlights the rational design of wide-ESW AHEs with high ionic activity as a promising approach to achieving portable and scalable applications of aqueous high-voltage Zn||MnO₂ batteries.

Rechargeable aqueous Zn||MnO₂ batteries have garnered substantial interest in recent years (1, 2). However, the practical applications of these batteries are greatly hindered by their modest discharging plateau (3) and limited areal capacity (4), mainly due to the single-electron redox reaction of MnO₂ (typically Mn⁴⁺↔Mn³⁺). In this regard, activating the multivalent conversion of MnO₂ (i.e., Mn⁴⁺↔Mn²⁺) is promising yet remains challenging. As shown in Fig. 1A, a discharging plateau at approximately 1.9 V can be obtained based on the MnO₂→Mn²⁺ conversion in a strong acidic electrolyte (consisting of 1 M ZnSO₄, 1 M MnSO₄, and 0.1 M H₂SO₄, details in SI Appendix, Fig. S1), the electrolyte will severely corrode the Zn anode, incurring low coulombic efficiency (CE) and undesirable side reactions (SI Appendix, Fig. S2A) (3, 5–7). When the water-in-salt (WIS) electrolyte was used, the distinct high discharging plateau could not be activated despite its acidic nature (Fig. 1B and SI Appendix, Fig. S2B) and extended electrochemical stability window (ESW) of 2.6 V (SI Appendix, Fig. S3A), possibly caused by the poor reactivity of MnO₂/Mn²⁺ couple as well as the low ionic activity (partially reflected by relatively low ionic conductivity of 0.52 S cm⁻¹ × 10⁻² of the electrolyte) (SI Appendix, Fig. S3B and Table S1). Of note, the MnO₂/Mn²⁺ redox reaction cannot be fully activated under constant current mode in the strong acid electrolyte and under constant voltage mode in WIS electrolytes (SI Appendix, Fig. S4). Such problems aggravate further when operating the electrolytic batteries at low rates (e.g., ≤1C) as required in realistic energy storage scenarios, and the alternating different charging/discharging modes (constant-voltage for charging and constant-current for discharging) and/or decoupled-electrolyte based batteries engender additional cost and complexity for the battery system.

Fig. 1.

Current drawbacks of electrolyte designs to activate high-voltage Zn||MnO₂ batteries and mechanism of our AHEs. (A) Charging-discharging curve of electrolytic Zn||MnO₂ battery (1 M ZnSO₄ + 1 M MnSO₄ + 0.1 M H₂SO₄). The Inset is a photograph of the Zn anode after cycling 20 cycles. (B) Charging-discharging curve of a full cell of Zn||MnO₂ battery assembled by a “WIS” electrolyte (12 M LITFSI + 1 M ZnSO₃CF₃ + 0.1 M MnSO₄). The Inset is a photograph of the Zn anode after cycling 20 cycles. (C) Schematic diagram of the hydrophobicity in the amphiphilic hydrogels enhancing the hydrogen bond strength. (D) Schematic diagram of the synthesis of AHEs.

Essentially, fulfilling the multivalent conversion of MnO₂ necessitates two prerequisites for the electrolyte: wide ESW allowing high valent conversion and a suitable reactive microenvironment that avails the reaction kinetics of the heterogeneous conversion of Mn species (i.e., Mn²⁺_(aq.)↔MnO_2(s)). Although superconcentrated electrolytes (e.g., “water-in-salt” electrolytes) (8) and/or additive-containing electrolytes (e.g., crowding agents) enable wide ESWs (9), the reactivity of the cations in these electrolytes is substantially compromised. Further, for high areal capacity rivaling that of existing Li-ion batteries (typically 3 to 4 mAh cm⁻²), effective utilization of the thick MnO₂ cathode layer is equally important, given the poor interfacial mass transport of Mn²⁺ poses another critical challenge. Hitherto, most existing electrolytes cannot meet all these criteria.

Hydrogel electrolytes have been scrutinized as an attractive catalog of electrolytes for Zn metal batteries, given their enriched interactions among polymer backbones, hydrated water molecules, and ions, enabling tunable electrochemical properties that meet customized battery requirements. Two mechanisms should be noted: Small purely hydrophobic solutes in water may increase the order and restrict the mobility of neighboring water molecules, and the resulting water hydrogen bonds in the neighborhood of hydrophobic moiety are as strong as those in ice clathrates (10–12). Therefore, by adding suitable amounts of hydrophobic moieties, the hydrogel electrolytes can enhance the hydrogen-bond interactions between hydrophilic groups and water molecules, leading to extended ESW even in an enriched water environment (Fig. 1C) (10–12). Then, the high ionic activity obtained in the hydrogel electrolytes indicates their potential for achieving high and flat discharging plateau in aqueous Zn||MnO₂ batteries (13, 14). In addition, the particular structure designability of the hydrogel electrolytes with brush-like spikes can optimize the cathode with high mass loading and control zinc deposition (002) to achieve high-rate aqueous Zn||MnO₂ batteries (15, 16). Developing amphiphilic hydrogel electrolytes (AHEs) with wide ESW and high-water content can help achieve scalable high-voltage aqueous Zn||MnO₂ batteries.

In this work, we synthesize the AHEs by a dynamic hydrophobic strategy (15–17), and the hydrogel electrolytes are equipped with extended ESW and high ionic activity. The designed hydrogel electrolytes consist of Mn²⁺ ions and crosslinked copolymer of hydrophilic acrylamide (AM), hexadecyl trimethyl ammonium chloride (CTAC), and hydrophobic stearyl methacrylate (C₁₈). The hydrogel electrolyte shows a wide ESW of ~3.0 V, high water content of ~76 wt% by mass percent composition, and high ionic conductivity of ~3.4 × 10⁻² S cm⁻¹. The assembled aqueous Zn||MnO₂ batteries exhibit a high-voltage and flat discharging plateau of ~1.9 V and a large areal capacity of ~5 mAh cm⁻² at 1 mA cm⁻² and still retain ~2.5 mAh cm⁻² even after cycling 320 h. When the mass loading is 1 mg cm⁻², the Zn||MnO₂ battery also shows an areal capacity of ~0.3 mAh cm⁻² after 10,000 cycles at a high current density of 3 mA cm⁻².

Results and Discussion

AHEs.

The AHEs are synthesized by employing a dynamic hydrophobic strategy (15–17). As illustrated in Fig. 1D, the AHE backbone consists of a crosslinked copolymer of hydrophilic AM and hydrophobic C₁₈ monomers, along with the sodium dodecyl sulfate (SDS) or CTAC surfactant. The surfactant was introduced to enable the micellar copolymerization of the monomers, which were covalently connected by a small quantity of chemical crosslinker. The as-prepared AHEs were denoted as PAM-SDS-C₁₈ or PAM-CTAC-C₁₈, respectively. The molecular structures of PAM-SDS-C₁₈ and PAM-CTAC-C₁₈ are exhibited in SI Appendix, Fig. S5. Such AHEs feature unique brush-like spikes for better ion migration and targeted adhesion properties (15–17). Unlike the acidic electrolytes, these AHEs are intrinsically neutral and mild (SI Appendix, Fig. S6), which is expected to render better stability and reversibility for the batteries.

Attenuated total reflection-Fourier-transform infrared (ATR-FTIR) spectroscopy identified the successful synthesis of these AHEs (SI Appendix, Fig. S7). In particular, the aliphatic carbon tails of all the fabricated AHEs were verified by two characteristic peaks (2,850 and 2,916 cm⁻¹), which are absent in the spectrum of the reference polyacrylamide (PAM) counterpart (SI Appendix, Fig. S7) (15–17). compared to PAM-SDS-C₁₈, PAM-CTAC-C₁₈ shows more apparent characteristic peaks of long aliphatic carbon tails (SI Appendix, Fig. S7). This result indicates that upon ionic strength solutions, the PAM-CTAC-C₁₈ AHE can better self-adjust into an interfacial hydrophobic structure than PAM-SDS-C₁₈ (15–17), potentially enabling interfacial adhesion and reactions (18).

Tuned Water States in the AHE.

Linear sweep voltammetry (LSV) profiles were recorded to determine the ESWs of the synthesized AHEs, with PAM and the aqueous acidic electrolyte solution (1 M ZnSO₄ + 0.1 M H₂SO₄) as references (Fig. 2A). Of note, owing to the conversion of Mn²⁺ to MnO₂ in cathodic voltage sweeping, electrolytes with Mn²⁺ will exhibit fake high potentials of oxygen evolution reaction (OER, >2.50 V vs. Zn²⁺/Zn of PAM and >4.50 V vs. Zn²⁺/Zn of PAM-CTAC-C₁₈ hydrogel electrolytes, SI Appendix, Fig. S8). The real potential of OER in PAM hydrogel electrolyte with Mn²⁺ can be identified by multiple scanning until the saturation of such conversion (that is, MnO₂ cannot be further deposited on the electrode surface). The real ESW of 1.90 V of the Mn²⁺-containing PAM hydrogel electrolyte is the same as that of the Mn²⁺-free PAM counterpart (Fig. 2A and SI Appendix, Fig. S8A). Accordingly, in order to completely circumvent the possible disturbance posed by Mn²⁺, the ESWs of various electrolytes were recorded based on Mn²⁺-free samples. Obviously, both AHEs display wider ESWs (2.9 to 3.0 V) than those of the control samples (1.65 to 1.90 V) (Fig. 2B). Notably, the precursor monomer concentration (~2 M) and the water content (~76 wt%) of the two AHEs are almost identical to that of PAM (SI Appendix, Fig. S9), yet the AHEs have an ionic conductivity, which is about 1.6 times higher than that of the PAM equivalent (SI Appendix, Fig. S10 and Table S1). Such distinctions between the ESW and ionic conductivity of the two types of gel electrolytes highlight that the introduced hydrophobic moieties substantially affect intermolecular hydrogen (H) bond interactions and ion transport. Of note, regardless of the specific hydrophilic or hydrophobic components utilized, prepared hydrogel electrolytes can achieve expanded ESWs as long as they align with the intended structural characteristics (SI Appendix, Fig. S11).

Fig. 2.

Water state in the amphiphilic hydrogel electrolyte. (A and B) The ESW (A) of aqueous electrolytes (scan rate of 5 mV/s) and (B) magnified views of the regions outlined near cathodic extremes in A. (C) TGA curves of PAM, PAM-SDS-C₁₈, and PAM-CTAC-C₁₈ hydrogel electrolytes at a heating rate of 5 °C/min. (D) DSC heating curves of PAM, PAM-SDS-C₁₈, and PAM-CTAC-C₁₈ hydrogel electrolytes at a heating rate of 10 °C/min. (E and F) Raman spectra of (E) PAM, PAM-SDS-C₁₈, and PAM-CTAC-C₁₈ hydrogel electrolytes and (F) O-H stretching vibration of PAM and PAM-CTAC-C₁₈ hydrogel electrolytes. (G) ATR-FTIR spectra of 0.1 M H₂SO₄ + 1 M ZnSO₄ + 1 M MnSO₄ solution electrolyte, PAM, PAM-SDS-C₁₈, and PAM-CTAC-C₁₈ hydrogel electrolytes. (H) Relaxation time (T₂) distribution curves recorded by LF-NMR of the synthesized PAM, PAM-SDS-C₁₈, and PAM-CTAC-C₁₈ hydrogel electrolytes.

Several essential characterizations were performed to reveal the underlying mechanism of the hydrophobic effect in the hydrogel electrolytes. Generally, the water species with hydrogel electrolytes can be classified into free water (FW), freezable bound water (FBW), and nonfreezable bound water (NFBW) (19–21). Thermogravimetric analysis (TGA) was first employed to explore the overall hydration states in different electrolytes. As shown in Fig. 2C, water volatilizes dramatically in the PAM hydrogel electrolyte (weight loss of 72.0 wt%) when the temperature (T) increases to 100 °C, indicating large amounts of FW that possesses weak interactions with the bare hydrophilic polymer chains (22). In contrast, a small weight loss of 14.0 wt% and 10.0 wt% was determined in the TGA profiles of PAM-SDS-C₁₈ and PAM-CTAC-C_18, respectively. After further increasing T to 200 °C, the 14.0 wt% weight loss that most probably results from the FBW species, was recorded in the TGA profile of the PAM-CTAC-C₁₈ hydrogel electrolyte. Given a high total water content of 76.2 wt% (SI Appendix, Fig. S9), the PAM-CTAC-C₁₈ hydrogel electrolyte has an NFBW content of 52.2 wt%, 1.72 and 21.8 times higher than those in PAM-SDS-C₁₈ and PAM gel electrolytes, respectively. Such stark contrast verifies that the hydrophobic effect can remarkably enhance the interactions between polymer chains and water molecules (10), and the PAM-CTAC-C₁₈ can better leverage this effect than PAM-SDS-C₁₈.

Differential scanning calorimetry (DSC) is another thermal analysis technique that can differentiate the three types of water in gel electrolytes (19–21). Fig. 2D shows that during the heating scan, the melting endotherm of PAM-CTAC-C₁₈ splits into two peaks at approximately −8 °C (peak 1) and −15 °C (peak 2), attributing to the FW and FBW, respectively. Though NFBW cannot be observed in the DSC profile, the much lower melting peaks of PAM-CTAC-C₁₈ than those of the other two control samples unambiguously validate the more intense water–polymer interactions (19–21).

In a microscale view, different water species are largely determined by the interwater H-bond. Given the oxygen-hydrogen (O-H) stretching mode (ν_O-H) is quite sensitive to the change of H-bond strength (8, 23), such features were then compared by Raman spectroscopy. Typically, the ν_O-H mode of H₂O at 3,200 to 3,400 cm⁻¹ in the AHEs exhibits an obvious blue shift to that of the PAM electrolyte (Fig. 2E). These shifts in the H-O covalent bond oscillation are induced by the perturbation of the water H-bond network from water–polymer interactions (9, 10). In other words, the interwater H-bond strength was weakened upon the introduction of the hydrophobic effect, probably because the hydrophobic polymer motif (C₁₈) repels water to its neighboring hydrophilic part (i.e., the amide group within acrylamide units) and reinforce water–polymer interactions. Further, the strongest characteristic bands (2,850 to 3,000 cm⁻¹) of long aliphatic carbon tails suggest the best stretching out of alkyl chains in the PAM-CTAC-C₁₈ electrolyte. Fig. 2F shows that the gel electrolytes can be best deconvoluted into three peaks centered at 3,240 cm⁻¹, 3,405 cm⁻¹, and 3,510 cm⁻¹ raised by the NFBW, FBW, and FW, respectively. Compared with PAM, PAM-CTAC-C₁₈ electrolyte presents a larger amount of NFBW and reduced FW, consistent with the TGA results. Similar conclusions were further supported by the ATR-FTIR spectroscopy results that compare the AHEs with PAM and aqueous acidic electrolyte solution (0.1 M H₂SO₄ + 1 M ZnSO₄ + 1 M MnSO₄) (Fig. 2G). In particular, the ν_O-H mode of H₂O at 3,200 to 3,400 cm⁻¹ in the AHEs exhibits an obvious blue shift compared with that in the PAM and aqueous solution.

We also performed low-field NMR (LF-NMR) spectroscopy to measure the transverse spin–spin relaxation time (T₂) that unravels the proton mobility in different gel electrolytes (19, 24). Detecting the attenuated signal resulting from the transverse spin–spin relaxation of protons via the Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence can probe the T₂ and the distribution of water molecules in the hydrogels, revealing their hydration behaviors (19). As such, the relaxation time distribution profiles of the three gel electrolytes are acquired from the CONTIN analysis based on the inverse Laplace transform on the CPMG decay curves. Compared to the PAM sample, the AHEs exhibit smaller T₂ (Fig. 2H). This trend corroborates again that water mobility is more substantially confined, and more bound water molecules exist in AHEs (particularly in PAM-CTAC-C₁₈), rendered by the stronger water–polymer interactions (19, 24). Taken together, all these combined thermal and spectrometry analyses substantiate that the hydrophobic effect can effectively diminish FW molecules within the gel electrolytes, even with a high total water content. As justified later, such desirable features could enable much wider ESWs while maintaining a favorable environment and a high ionic activity/ionic conductivity for the conversion reactions.

Molecular Insights of Water/Ion–Polymer Interactions.

To review the molecular scale influence of the hydrophobic effect within the AHEs, molecular dynamics (MD) simulations were performed (25–37). The hydrophobic effect was examined by evenly incorporating different amounts of C₁₈ moiety (0, 2, 4, or 6 units) into the PAM main chains (each consisting of 32 units). Other parameters are detailed in SI Appendix, Table S2. Fig. 3 A–D show the snapshots of the relaxed gel electrolyte boxes. It is noted that with the introduction of a suitable amount (2 or 4 C₁₈ units per PAM chain) of hydrophobic moieties, the H-bond within the hydrated polymer backbone was strengthened, with an average PAM–water interaction energy increased from the initial 12.5 to 14.5 kcal mol⁻¹ (Fig. 3E). This trend suggests that, due to the localized hydrophobic effect, the neighboring water molecules are more intensely confined by (or being pushed toward) the hydrophilic part of the AHEs. Further increasing the C₁₈ concentration will lead to a declined interaction energy, possibly because an excess amount of C₁₈ will agglomerate by the hydrophobic association, in line with the experimental tests. It should be noted that all the molecules and ions are fully relaxed in the aqueous environment; other factors, such as crosslinking and realistic cohesion force within the hydrogel that could further contribute to the strengthened water–polymer interactions, are not considered here. In other words, such interactions are somehow underestimated in the present MD models. Hence, the MD simulation elucidates that the hydrophobic effect-induced water confinement can indeed expand ESW within the AHE, even in a water-enriched environment.

Fig. 3.

Simulation of the hydration analyses. (A–D) MD of water molecules in PAM-CTAC-C₁₈ hydrogel electrolytes with different amounts of C₁₈ (0, 2, 4, or 6 units) into the PAM main chains (each consisting of 32 units). (A) 0. (B) 2. (C) 4. (D) 6. (E) H₂O–hydrogel interaction energy in various PAM-CTAC-C₁₈ hydrogel electrolytes. (F) The radial distribution function and CN of Mn ion and water molecules [dashed: n(r)]. (G) The radial distribution function and CN of Mn ion and AM molecules [dashed: n(r)]. (H) The desolvation energy of Mn ion in the PAM-CTAC-C₁₈ hydrogel electrolyte and solution electrolyte.

Apart from this, the localized hydrophobic effect could also affect the ion–polymer interactions. As such, the solvation-sheath structures of the Mn²⁺ in AHEs with different C₁₈ content were revealed by the radial distribution function and corresponding coordinated numbers (CNs). The radial distribution function profiles (Fig. 3F) show that the water molecules entered the first solvation sheath of Mn²⁺ in all hydrogel electrolytes. However, the optimal localized hydrophobic effect repels the aqua ligand away, with the 2-C₁₈ units per PAM chain leading to a notably decreased water CN. Meanwhile, it is interesting to note that the coordination of Mn²⁺ with amine groups in AM moieties (Fig. 3G) was enhanced with increased CNs in Mn-O (AM). Such trade-off in CNs of water molecules and amine groups (from hydrophilic PAM chains) in the Mn²⁺ first solvation sheath indicates the introduced hydrophobic effect in AHEs can essentially regulate the ion–polymer interactions in the context of the cation solvation structure. As a result, it can be envisaged that the desolvated Mn²⁺ could have distinct reaction kinetics in subsequent conversions. Hence, we carried out density functional theory (DFT) calculations to compare the desolvation energy of the [Mn(AM)(H₂O)₅]²⁺ and [Mn(H₂O)₆]²⁺ clusters (the detailed calculation process can be seen in SI Appendix). Fig. 3H and SI Appendix, Fig. S12 exhibit that all the stepwise desolvation energies of [Mn(AM)(H₂O)_n]²⁺ (n = 6-1) clusters are much lower than those of their [Mn(H₂O)_n]²⁺ counterparts. For example, the energy requested by [Mn(AM)(H₂O)₅]²⁺ to remove the first water molecule is approximately half of that done by [Mn(H₂O)₆]²⁺. Therefore, with an external voltage employed, [Mn(AM)(H₂O)₅]²⁺ is more likely to remove H₂O molecules than that of [Mn(H₂O)₆]²⁺. This facilitated desolvation process of hydrated Mn²⁺ will circumvent the competing OER and decrease its nucleation overpotential (NOP) (18). Of note, the solvation energy barriers of Mn²⁺ are positively correlated with the desolvation energy barriers of hydrated Mn²⁺ in AHEs and (1 M ZnSO₄ + 0.2 M MnSO₄) solution electrolyte (SI Appendix, Fig. S13). The combined DFT and MD results thus imply that the Mn²⁺↔MnO₂ conversion reaction could occur more easily on the surface of the cathode with the assistance of the wide-ESW-AHE, whereby the hydrophilic parts boost the desolvation steps and the hydrophobic moieties repel water and lubricate the transport of Mn²⁺.

The AHE-Enabled High-Voltage Zn||MnO₂ Batteries.

Given the above-desired features, the PAM-CTAC-C₁₈ AHE was employed as an electrolyte for conversion-type Zn||MnO₂ batteries, with a normal PAM gel electrolyte as a reference. As shown in Fig. 4A, the starting NOP decreases from 1.84 to 1.68 V with the introduction of C₁₈ in the PAM hydrogel electrolyte, indicating that the PAM-CTAC-C₁₈ AHE is indeed conducive to promoting Mn²⁺ nucleation, consistent with the theoretical calculations (Fig. 3). As such, the batteries were first evaluated per the previous protocol: first, a constant-voltage charge at 2.2 V, followed by a typical galvanostatic discharge at 0.5 mA cm⁻². Fig. 4 B and C display the profiles of the Zn||MnO₂ batteries with different electrolytes. The MnO₂/Mn²⁺ redox conversion cannot be completely activated with PAM hydrogel electrolyte, yielding a low and fluctuating CE even below 40% (Fig. 4B). By contrast, using PAM-CTAC-C₁₈ AHE, such a reaction was successfully leveraged, and a high discharge plateau at 1.9 V can be realized (Fig. 4C). More significantly, after an initial few cycles, this preferential conversion-based charge-discharge cycle can be well sustained for 100 cycles with an average CE approaching 100% (Fig. 4D). Such stark distinctions in the assembled batteries corroborate the AHE’s innate wide ESW and facilitated interfacial kinetics can boost the conversion reaction stably. Of note, the voltage profiles based on PAM-CTAC-C₁₈ AHE at the 50th and 100th cycles under constant voltage charge mode (Fig. 4C) demonstrate the distinct MnO₂/Mn²⁺ redox reaction at a high potential of >1.8 V vs. Zn²⁺/Zn. The slight voltage drop in the high-voltage discharge plateau (>1.8 V) of the aqueous Zn||MnO₂ battery can be attributed to a gradual increase in impedance after 100 cycles (SI Appendix, Fig. S14A), which is consistent with the performance of the beaker battery (using an exceptionally thick Zn foil) assembled with a strong acid electrolyte (SI Appendix, Fig. S14B). In addition, the low-voltage discharge plateaus at 1.3 and 0.8 V are attributed to Zn²⁺/H⁺ intercalation into residual MnO₂, which is induced by two processes: the incomplete desolvation of hydrated Mn²⁺ during the electrochemical process and the gradually generated (de)intercalation-type MnO₂ crystal particles on the electrodes (SI Appendix, Figs. S15 and S16). However, it is worth noting that a charging voltage that is too high (e.g., 2.4 V or 2.6 V) can destroy the crystal structure of MnO₂, leading to irreversible conversion and compromised CEs (SI Appendix, Fig. S17).

Fig. 4.

Electrochemical behaviors of high-voltage Zn||MnO₂ batteries. (A) LSV curves in PAM and PAM-CTAC-C₁₈ hydrogel electrolytes. Galvanostatic discharge curves (0.5 mA cm⁻²) at the constant voltage of 2.2 V in different electrolytes. (B) PAM (1 M ZnSO₄ + 0.2 M MnSO₄) (C) PAM-CTAC-C₁₈ (1 M ZnSO₄ + 0.2 M MnSO₄). (D) The corresponding cycling stability test of (c) at 0.5 mA cm⁻². (E) Charging-discharging curves and (F) Cycle performance of pouch cells assembled by PAM-CTAC-C₁₈ hydrogel electrolytes with a loading mass of MnO₂ of 0 mg cm⁻² (current density of 2 mA cm⁻²).

Afterward, the batteries were scrutinized under the more challenging constant-current cycling mode. At first, the galvanostatic charge-discharge (GCD) profiles and cycle performance of the PAM-CTAC-C₁₈ based battery with bare carbon cloth (CC) current collector (i.e., 0 mg cm⁻² of cathode loading mass) were recorded. Here again, the MnO₂/Mn²⁺ conversion-induced discharge plateau at 1.9 V (vs. Zn²⁺/Zn) can be activated (Fig. 4E), and the battery retained excellent stability and reversibility during the following 5,000 cycles (totaling ~1,200 h) (Fig. 4F). The cyclic voltammetry (CV) curve also clearly shows a reduction peak at about 1.9 V arising from the MnO₂→Mn²⁺ conversion in the PAM-CTAC-C₁₈ based battery, in contrast to the control battery assembled from PAM-based electrolyte (SI Appendix, Fig. S18). Differential electrochemical mass spectrometry (DEMS, SI Appendix, Fig. S19) is conducted to reveal that OER is not the sole factor for activating Mn²⁺/MnO₂ conversion reaction, though, at constant 2.2 V, the OER is more significantly activated compared to the constant current mode. Note that the other AHE, PAM-SDS-C_18, based batteries also present prominent redox peaks of MnO₂/Mn²⁺ in the CV profiles, though their GCD discharge plateau is lower than those with PAM-CTAC-C₁₈ (SI Appendix, Fig. S20). The distinction can be attributed to the strong complexing effect between -SO₄⁻ groups in SDS and Mn²⁺, resulting in relatively lower Mn²⁺ activity and terrible reactive microenvironment. These control experiments manifest that the overall hydrophobic effect with the AHEs is general, and the more intense water confinement in PAM-CTAC-C₁₈ (Fig. 2 C–H) presumably accounts for the batteries’ better electrochemical performances. Such a trend was confirmed to be solid with batteries scanned with other voltage ranges and different current collectors using this constant-current cycling mode (SI Appendix, Figs. S21–S23). It should also be noted that excess amounts of hydrophobic moieties (C₁₈) in PAM-CTAC-C₁₈ AHE will lead to a lower CE of <90% in the assembled Zn||MnO₂ battery (SI Appendix, Fig. S24). This is understandable since too many hydrophobic moieties in the amphiphilic hydrogel could cause phase segregation, which breaks the balanced water- and ion–polymer interactions. Such results also align well with the MD results (Fig. 3 A–G).

The conversion process of the cathode was also then investigated with an ex-situ X-ray diffraction (XRD) analysis. SI Appendix, Fig. S25 displays a typical GCD profile of the PAM-CTAC-C₁₈ based battery and the XRD patterns of the cathode (preloaded MnO₂-CC film) recorded at various voltage stages. As the battery discharges from 2.15 to 0.83 V, the intensity of characteristic peaks of MnO₂ (JCPDS 04-007-8867) decreases gradually (SI Appendix, Fig. S25B), due to the dissolution of MnO₂. Notably, at deep discharge voltages of 1.16 and 0.83 V, the characteristic peaks of MnOOH (JCPDS 18-0805), ZnMn₂O₄ (JCPDS 19-1459), and Zn₄SO₄(OH)₆·4H₂O (JCPDS 44-0675) emerge (3, 7). In contrast, during charging from 1.20 to 2.20 V, the width of the characteristic peaks of MnO₂ broadens gradually, which is reasonably caused by the accumulation of sedimentary MnO₂ species that differ from the precoated α-MnO₂ phase (38). Besides, when the battery is charged to 2.20 V, the disappearance of characteristic peaks of ZnMn₂O₄ and Zn₄SO₄(OH)₆·4H₂O indicates the desertion of Zn²⁺ and H⁺ ions (SI Appendix, Fig. S25C). These observations suggest that the whole discharge/charge process comprises a mixed dissolution-intercalation mechanism, during which MnO₂ was first dissolved/deposited, and then the intercalation/extraction of Zn²⁺ and H⁺ ions occurs. These results also elucidate why the GCD profile of the AHE-based battery features a distinct high-voltage discharge plateau followed by a typical ion-insertion one during discharge. SI Appendix, Fig. S26A shows that the charge transfer resistance of the AHE-based battery is substantially lower than that of the PAM battery, even resembling the case in the electrolytic Zn||MnO₂ battery using strong acid electrolyte (SI Appendix, Fig. S26B). Of note, the impedance spectra of three-electrode cells assembled with these electrolytes were conducted to eliminate the interference from the Zn anode (SI Appendix, Fig. S27), and the results are consistent with the trends of the impedance results of full Zn||MnO₂ batteries. These results consolidate that, despite the charging-discharging modes, the AHE can essentially enable the high-voltage conversion of MnO₂↔Mn²⁺ in Zn||MnO₂ batteries, featuring kinetics comparable with that in conventional strong acid electrolyte but with much-enhanced stability and reversibility.

High-Areal Capacity and Scalable Zn||MnO₂ Batteries.

In contrast, we also detailed the electrochemical performance of PAM-SDS-C₁₈ (SI Appendix, Fig. S28) and PAM-CTAC-C₁₈ batteries after precoating MnO₂ on the surface of the CC film. First, when the loading mass is 1.0 mg cm^−2, and the current density is 3 mA cm⁻², the PAM-CTAC-C₁₈ battery exhibited 10,000 cycles and a mean areal capacity of 0.3 mAh cm⁻² all the time (Fig. 5 A and B). SI Appendix, Fig. S29 further identified the excellent cycle performance of PAM-CTAC-C₁₈ batteries. Fig. 5 C–E revealed the high loading mass of 15.6 mg cm⁻² and large areal capacity of close 5 mAh cm⁻² while a long cycling time of >320 h (Fig. 5E). The performances of the full Zn||MnO₂ batteries assembled with the PAM-CTAC-C₁₈ AHE were compared with previously reported aqueous Zn||MnO₂ batteries (Fig. 5F) (3, 7, 18, 39–42). The developed Zn||MnO₂ full batteries exhibited the best performance in terms of a high discharge plateau of MnO₂/Mn²⁺ conversion and areal capacity.

Fig. 5.

Cycling performance of high-voltage Zn||MnO₂ pouch cells. (A) Charging-discharging curves and (B) Cycle performance of pouch cells assembled by PAM-CTAC-C₁₈ hydrogel electrolytes when loading mass of MnO₂ is 1 mg cm⁻² (current density of 3 mA cm⁻²). (C) Cycle performance of pouch cells assembled by PAM-CTAC-C₁₈ hydrogel electrolytes of loading different masses of MnO₂ on the current collector (current density of 1 mA cm⁻²). (D) Charging-discharging curve of loading mass of 15.6 mg cm⁻² and (E) the corresponding cycling performance. (F) Comparison of the discharge plateaus of MnO₂/Mn²⁺ conversion and areal capacities of the full batteries with reported aqueous Zn||MnO₂ batteries. (G) Charging-discharging curves of scalable (40 cm²) Zn||MnO₂ pouch cell. The shelf life of (H) the pouch cell assembled by PAM-CTAC-C₁₈ (incorporating 1 M ZnSO₄ + 0.2 M MnSO₄) and (I) the coin cell assembled by 1 M ZnSO₄ + 1 M MnSO₄ + 0.1 M H₂SO₄.

According to our previous research, the brush-like spikes of AHEs showed the targeted adhesion on the cathode, resulting in high stability and full use of the cathode materials (SI Appendix, Fig. S30 A and C). On the other hand, it can also regulate the Zn deposition (002) (15, 43) (SI Appendix, Figs. S30 B and D and S31) to achieve highly reversible Zn deposition and stripping; in contrast, the strong acid electrolytes previously used to achieve MnO₂ deposition/dissolution will erode Zn anode (SI Appendix, Fig. S30B). In detail, the Zn||Zn symmetric cells assembled by PAM-CTAC-C₁₈ hydrogel electrolyte exhibited long cycle life at 3 mA cm⁻² and 3 mAh cm⁻² (SI Appendix, Fig. S32 A and B). However, the strong acid and (12 M LiTFSI + 1 M Zn(SO₃CF₃)₂ + 0.1 M MnSO₄) solution electrolytes showed poor Zn deposition and stripping (SI Appendix, Fig. S32C).

Therefore, targeted adhesion to the cathode and regulation to the Zn deposition (002) provide the potential for fabricating large-scale aqueous Zn||MnO₂ batteries with excellent electrochemical performance. In particular, we assembled a pouch cell with 40 cm², achieving the capacity of 125 mAh at 2 mA cm⁻² (Fig. 5G). Of note is that the relatively inconspicuously 1.9 V discharging plateau could be attributed to its too-high capacity that widely appears in many aqueous battery systems (4, 44), which obscured the emergence of the plateau due to the increasing diffusion resistance of ions at the cathode. On the other hand, extremely high mass loading, large area, and current density will cause the increasing diffusion resistance of the ions. Then, the ionic activity is reduced. Subsequently, a few high-voltage Zn||MnO₂ batteries with 25 cm² were fabricated. Two pouch cells connected in series can drive a 3.7 V-powerable electric fan that is supposed to be driven by organic lithium-ion batteries (SI Appendix, Fig. S33 and Movie S1). The output power of the electric fan is 1.5 W, and other parameters can be seen in SI Appendix, Fig. S33A. This clearly revealed the scalability of large-scale aqueous Zn||MnO₂ batteries with high discharging plateau.

Finally, we assess the shelf life of PAM-CTAC-C₁₈ based batteries and Zn||MnO₂ batteries assembled by strong acid solution electrolyte (1 M ZnSO₄ + 1 M MnSO₄ + 0.1 M H₂SO₄). As shown in Fig. 5H, after resting for 24 h for 5 times, PAM-CTAC-C₁₈ based batteries exhibited no capacity fade-off. In contrast, Zn||MnO₂ coin cell assembled by strong acid solution electrolyte (1 M ZnSO₄ + 1 M MnSO₄ + 0.1 M H₂SO₄) first showed low CE of <50% and resting for 24 h at the third time, and then the coin cell completely destroyed (Fig. 5I). Although the beaker cell assembled by strong acid solution electrolyte (1 M ZnSO₄ + 1 M MnSO₄ + 0.1 M H₂SO₄) showed relatively good electrochemical performance compared to the coin cell after resting for 24 h for 5 times, the beaker battery was also completely destroyed owing to the erosion of Zn foil (SI Appendix, Fig. S34). Therefore, PAM-CTAC-C₁₈ hydrogel electrolytes with a mild environment and expanded ESW possess remarkable potential in scalable high-voltage Zn||MnO₂ batteries.

In summary, AHEs with extended ESWs and high ionic activity are fabricated and employed in rechargeable Zn||MnO₂ batteries. Trace amounts of hydrophobic moieties can enhance hydrogen bonding between hydrophilic groups and water molecules, enabling a wide ESW and activating the MnO₂/Mn²⁺ redox conversion. Meanwhile, the high ionic activity of the AHEs stabilizes the MnO₂/Mn²⁺ redox conversion, resulting in a high-voltage and flat discharging plateau of the Zn||MnO₂ batteries. The developed AHE possesses an ESW up to ~3.0 V even at a high water content of ~76 wt%. On the other hand, the unique structure of AHEs endows Zn||MnO₂ batteries to take full use of the thick cathode and regulate the (002) zinc deposition, resulting in the large areal capacity and long cycles of the Zn||MnO₂ batteries. In addition, the assembled large-scale Zn||MnO₂ batteries exhibit scalable performance. Particularly, a 40 cm² pouch cell demonstrates the capacity of 125 mAh at 2 mA cm⁻². Meanwhile, two 25 cm² pouch cells connected in series drive a 3.7 V-powerable electric fan. This exquisite designability creates a chance to obtain aqueous Zn||MnO₂ batteries with high-voltage and flat discharging plateau at 1.9 V and a large areal capacity of 5 mAh cm⁻². This opens a broad avenue to prompt the development and practical application of large-scale Zn||MnO₂ batteries.

Materials and Methods

Materials.

Acrylamide (AM), 2-acrylamido-2-methylpropanesulfonic acid (AMA), initiator ammonium persulfate (APS), N, N’-methylenebisacrylamide (MBAA), sodium dodecyl sulfate (SDS), hexadecyl trimethyl ammonium chloridestearyl (CTAC), methacrylate (C₁₈), trifluoroethyl methacrylate (3FMA), manganese sulfate monohydrate (MnSO₄·H₂O) and zinc sulfate heptahydrate (ZnSO₄·7H₂O), lithium bistrifluoromethanesulfonimidate (LiTFSI), lithium trifluoromethanesulfonate (LiSO₃CF₃), and manganese chloride (MnCl₂) were purchased from Aladdin and used directly. Zinc chloride (ZnCl₂) was purchased from Tixiai (Shanghai) Chemical Trading Co., Ltd. and used directly. Deionized (DI) water was used in all the experiments.

Synthesis of AHEs.

The AHEs were fabricated following the reference (15–17). In a typical run, SDS or CTAC (2.12 g) was first dissolved in 1 M ZnSO₄ solution (30 mL) at 100 °C. Second, 0.26 g C₁₈ (or 3FMA) was added to the solution, and the mixture was stirred at 100 °C for 10 min. Third, 1.5 g AM (or AMA) monomers were dissolved into the prepared solution. After 10 min of stirring, 0.003 g MBAA as crosslinker was added into the mixture, and then the solution was cooled to room temperature so that it would not cause copolymerization when 0.08 g APS was mixed. The mixture was then transferred to a customized mold. Finally, the sealed mold was placed in a vacuum oven at 70 °C for 1 h, and the hydrogels were obtained. The as-prepared hydrogels were immersed into the solution (1 M ZnSO₄ + 0.2 M MnSO₄) for a period, and the AHEs were obtained. We denoted AHEs as PAM-SDS-C₁₈, PAM-CTAC-C₁₈, PAM-CTAC-3FMA, and PAMA-CTAC-C₁₈ based on using different surfactants, hydrophilic and hydrophobic monomers.

Synthesis of PAM Hydrogel Electrolyte.

Radical polymerization was employed to fabricate the PAM hydrogel. Typically, 2.84 g AM and 0.002 g MBAA were dissolving into 20 mL DI water. Then, a clear solution can be obtained by magnetically stirring. Next, we transfer the mixture to the customized mold and place it in a vacuum oven for 60 min. After the reaction, PAM hydrogel was fabricated. Finally, the as-prepared PAM hydrogel was immersed into the solution (1 M ZnSO₄ + 0.2 M MnSO₄) for a period, and PAM hydrogel electrolyte was obtained.

Preparation of α-MnO₂ Cathode.

The α-MnO₂ powders were synthesized via a hydrothermal method (15). First, 2 mL 0.5 M H₂SO₄ and 0.507g MnSO₄·H₂O were added into 90 mL DI water. Second, 20 mL 0.1 M KMnO₄ solution was slowly added to the mixture. The obtained red mixture was stirred for 2 h, poured into the Teflon-lined autoclaves, and heated at 120 °C for 12 h. Next, DI water was used to wash the precipitates, and the mixture was centrifuged three times. Finally, the precipitates were placed in a vacuum oven for 24 h at 30 °C, and the α-MnO₂ powders were obtained.

To fabricate α-MnO₂ cathode, α-MnO₂ powder, ketjenblack (Sigma), and polyvinylidene fluoride (Sigma) were mixed with a weight ratio of 7:2:1, and the mixture was vigorously blended and ground. The N-methyl pyrrolidone (Aladdin) solvent was then added to the mixtures. Next, a homogeneous mixture slurry was obtained by stirring. Finally, the slurry was coated on the surface of a CC (Suzhou Tan Feng Technology Co., Ltd.). After drying them in a vacuum oven at 60 °C for 10 h, the α-MnO₂/CC cathode was fabricated.

Determine Water Content in Hydrogel Electrolytes.

Actual water contents of hydrogel electrolytes were calculated following the equation below:

$$\mathrm{Water} \mathrm{content} (wt\%)=(W_w-W_d)\ast 100\%/W_w,$$

Where W_w represents the weight of the wet hydrogel electrolyte, and W_d represents the corresponding weight of the dried hydrogel electrolyte. Of note, three samples for each hydrogel electrolyte were tested and averaged. The detailed calculation results are shown in SI Appendix, Fig. S8.

Electrochemical Measurements.

CV profiles were recorded by employing the electrochemical workstation (CHI 760e). The LAND test system was used to test the cycle performance. The voltage–time profiles of Zn||Zn symmetric cells were tested at the current densities of 1 or 3 mA cm⁻² with the areal capacity of 1 or 3 mAh cm⁻² (thickness of Zn foil is 50 μm). All tests were performed with pouch cells.

Simulation.

MD simulations are performed by the large-scale atomic/molecular parallel simulator package (25). The TIP3P water model (26) is used. Force field parameters for Zn²⁺, Cl⁻ are from Merz’s work (27, 28), and those for SO₄²⁻, CTA⁺, C₁₈, and AM are from the general Amber force field (GAFF) (29). The atomic partial charge coupled with GAFF is calculated by RESP approach (30) using the Multiwfn software (31) with the molecule structure optimized by ORCA (32) under the level of B3LYP/6-311G**. For mixing the parameters, the Lorentz-Berthelot rule is used. A cutoff distance of 10 Å is set for nonbonded interaction, and the long-range coulombic interaction is computed by the particle–particle particle–mesh method (33). The Nose/Hoover thermostat and barostat are used to control the temperature and pressure, respectively (34).

The Moltemplate program completes the construction of the initial structure for simulation (35). The chain length of 32 AM is chosen, which is enough to converge the solubility parameter of PAM (36). Four electrolyte systems are designed to evaluate the effect of C₁₈ and CTAC on the water binding with PAMs. In these systems, different numbers of C₁₈ are engaged in the polymerization by putting C₁₈s in positions that split the PAM chain as equally as possible. Molecule numbers packed in the simulation box for different electrolyte systems are listed in SI Appendix, Table S2. To reach equilibrium, the system is first heated with an NVT ensemble of 500 K for 2 ns and cooled down to 300 K with an NPT ensemble for 3 ns, followed by a 30 ns NPT run of 300 K. Then, 2 ns production run with NVE ensemble is used for analysis. The interaction energy between AM molecules and water is calculated every 2 ps, and the trajectory for H-bond analysis is saved every 2 ps. The visualization is realized by the VMD software (37).

DFT Calculations.

Both the geometry optimization and energy calculation are performed by ORCA (32) under the level of B3LYP/6-311+G**. To evaluate the degree of difficulty for the desolvation of the Mn solvation structure with and without AM, we define the successive desolvation energy of Mn[H₂O]_n[AM]²⁺ and Mn[H₂O]_n[H₂O]²⁺ (n is an integral from 1 to 5) as

$$\mathrm{\Delta }E1=E\left(\text{Mn}{\left[{\mathrm{H}}_2\mathrm{O}\right]}_{\mathrm{n}-1}{\left[\text{AM}\right]}^{2+}\right)+E({\mathrm{H}}_2\mathrm{O})-E\left(\text{Mn}{\left[{\mathrm{H}}_2\mathrm{O}\right]}_{\mathrm{n}}{\left[\text{AM}\right]}^{2+}\right),$$

and

$$\mathrm{\Delta }E2=E\left(\text{Mn}[{\mathrm{H}}_2\mathrm{O}{]}_{\mathrm{n}-1}[{\mathrm{H}}_2\mathrm{O}{]}^{2+}\right)+E({\mathrm{H}}_2\mathrm{O})-E\left(\text{Mn}{\left[{\mathrm{H}}_2\mathrm{O}\right]}_{\mathrm{n}}{\left[{\mathrm{H}}_2\mathrm{O}\right]}^{2+}\right),$$

where $\begin{array}{c}E\left(\mathrm{Mn}{\left[{\mathrm{H}}_2\mathrm{O}\right]}_{\mathrm{n}-1}{\left[\mathrm{AM}\right]}^{2+}\right), E\left(\mathrm{Mn}{\left[{\mathrm{H}}_2\mathrm{O}\right]}_{\mathrm{n}}{\left[\mathrm{AM}\right]}^{2+}\right), E\left(\mathrm{Mn}[{\mathrm{H}}_2\mathrm{O}{]}_{\mathrm{n}-1}\right.\hfill \end{array}$$\begin{array}{c}\left.[{\mathrm{H}}_2\mathrm{O}{]}^{2+}\right)\hfill \end{array}$, and are the total energies of the corresponding optimized complex, and $E({\mathrm{H}}_2\mathrm{O})$ is the total energy of a water molecule. The successive desolvation energy refers to the energy needed to extract one water molecule from the previous solvation structure.

Characterizations.

An environmental scanning electron microscope (FEI/Philips XL30) was employed to measure the surface morphology of Zn deposition. FTIR spectroscopy (PerkinElmer) profiles and Raman spectroscopy (WITec RAMAN alpha 300R) were carried out to separately measure the structure characteristics of the hydrogels and water molecules state in the hydrogels. XRD (D2 PHASER) patterns were recorded to identify the characteristics of the crystal particles. LF-NMR was performed using MesoMR23-060H-I (Niumag). TGA (TA Discovery TGA 550) was carried out ranging from 30 to 200 °C at 5 °C/min. DSC (25, TA) was employed to record the enthalpy change of hydrogel electrolytes when changing temperatures. The samples were measured with the following procedures: First, the sample would be cooled from 20 to −50 °C at a cooling rate of 10 °C/min and then maintained for 3 min. Next, it would be heated to 20 °C at a heating rate of 10 °C/min and finally held at 20 °C for around 3 min. The loop runs two times. The gas evolution was checked in a pouch cell (4.4 cm × 3.2 cm, the test module size) of an in-situ DEMS set-up (Shanghai Linglu Instrument Equipment).

Supplementary Material

Appendix 01 (PDF)

Movie S1.

Video demonstration of two pouch cells connected in series for driving a 3.7 V-powerable electric fan.

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information.