Aqueous Supercapacitor with Ultrahigh Voltage Window Beyond 2.0 Volt

Supercapacitors (SCs), a new type of green electrochemical energy storage device with high power density and long‐term durability, show great potential to replace nonrenewable energy sources. Benefiting from both economic and environmental advantages, aqueous SCs represent a fabulous prospect in many industries. However, the relative low voltage window of aqueous SCs hinders their further development. In the recent years, especially within the last five years, the significant achievements in aqueous SCs with ultrahigh voltage window (>2.0 V) have been reported. Herein, the effects of theoretical mechanisms on voltage window are first introduced, which provide the fundamental guidance to enlarge the voltage window of aqueous SCs. Subsequently, the strategies for constructing the aqueous SCs over 2.0 V are comprehensively summarized and classified into the electrode modification by structural engineering, metal cations doping and constructing advanced composites, and the electrolyte optimization by preparing “Water in Salt” and novel mixed electrolyte. Finally, via the discussion of current progresses and drawbacks of these >2.0 V aqueous SCs, their future development directions are proposed.

appealing for large-scale production ( Figure 1). [15] Unfortunately, the voltage window of aqueous SCs is confined in a narrow range. Thermodynamically, the electrochemical stability window (ESW) of aqueous electrolytes is 1.23 V, which is caused by undesirable water decomposition. [16] During the electrolysis of water solvent, the electrode-electrolyte interface will undergo hydrogen/oxygen gas evolution and subsequent concentration change of aqueous electrolyte. These phenomenon would significantly damage the stability of electrode-electrolyte interface and SCs electrochemical performance. [1d] Accordingly, to enhance voltage window of aqueous SCs without sacrificing cycle lifespan and power density, the adjustment for kinetic effects of electrolyte and other relevant mechanisms have to be introduced via the optimization of electrode/electrolyte materials. [16a] In the recent studies, particularly from the year of 2016, it is inspiringly surprising to discover many aqueous SCs with excellent electrochemical performance can not only push the voltage window upon the theoretical water-splitting potential window of 1.23 V but also achieve the ultrahigh voltage window that exceed 2.0 V, which reveals their huge potentialities in the future EES devices market.
To the best of our knowledge, although there have been limited reviews concerning aqueous SCs specialized in metal oxide and hydroxide-based materials, [17] and ones collecting representative surface engineering techniques, [18] and influenced factors for operating voltage windows, [19] a review is urgently needed that systematically and exclusively summarizes the reported aqueous SC illustrations with ultrahigh voltage window of over 2.0 V, and focuses on their novel strategies to achieve this goal. We believe the integration of these up-to-date achievements can provide the comprehensive inspiration for exploring high energy density aqueous SCs with stable output.
Herein, we thoroughly summarize the state-of-the-art progress of the aqueous SCs with ultrahigh voltage window (>2.0 V). The review first introduces the theoretical mechanisms which influence SCs voltage window, including that of aqueous electrolyte electrolysis and work function. Afterward, the advanced strategies of the >2.0 V aqueous SCs are systematically classified into two major categories: electrode modification and electrolyte optimization, in which the former specifically contains structural engineering, metal cations doping and constructing advanced composites, and the later includes using "Water in Salt" (WIS) and novel mixed electrolyte. The mechanisms of each strategy are analyzed, along with discussion of their advantages and disadvantages for large-scale practical applications. Finally, on account of the advancements and challenges of aforementioned aqueous SCs, we propose the promising development routes for aqueous SCs with ultrahigh voltage window in future.

Influence of Theoretical Mechanism on Voltage Window of SCs
The kinetic effects of electrolyte electrolysis play a vital role in aqueous SCs. The voltage window can breakthrough water thermodynamic stability region of 1.23 V by decreasing the activity of hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) (i.e., boosting HER/OER overpotentials), in which the dynamic mechanism can be realized via optimizing the electrode-electrolyte interaction. According to generally accepted theoretical mechanism, HER occurs on the negative electrode via electrochemical proton adsorption (Volmer reaction), followed by either an electrochemical desorption (Heyrovsky reaction) or chemical desorption (Tafel reaction) to generate gaseous hydrogen. [20] Similar with HER mechanism, OER is also a multi-step electrochemical reaction to produce oxygen gas based on adsorption of OH À onto positive electrode. Thus, reducing the ease of adsorption and removal of H þ /OH À can effectively raise HER/ OER overpotentials resulting in an incremental electrode potential.
Except HER/OER mechanism, other relevant mechanisms also contribute to the widening of SCs voltage window. Despite most of those mechanisms have not been clearly revealed, the work function has been proven to display intimate correlation with voltage window in asymmetric SCs (ASCs) (Figure 2a). It is evident that higher work function difference (WFD) between anode and cathode can provide wider voltage window. [21] The formulation of estimated voltage window (E) for SCs is shown below [23] where E 0 represents the Galvani potential associated with the work function of metal oxide/metal electrode; ΔE 1 and ΔE 2 denote surface dipole potentials of cathode and anode, respectively, which are relative to the overpotentials of OER and HER; ω α and ω β are positive and negative work functions, respectively; N A and F are Avogadro and Faraday constant, respectively. Hence, the WFDs between two electrodes are one of the determined factors to enlarge the voltage window in ASCs devices, especially with the known work function (Figure 2b). Nevertheless, the voltage window of pure metal oxides-based www.advancedsciencenews.com www.small-structures.com ASCs can hardly exceed 2.0 V in aqueous electrolyte. For instance, Chang et al. [22] fabricated the ASCs by using MnO 2 as positive electrode and MoO 3 as negative electrode due to their largest WFD among various metal oxides (4.4 and 6.2 eV, respectively), which provided the working voltage of 2.0 V (Figure 2c-e). For symmetric SCs (SSCs), the voltage window is almost determined by ΔE 1 and ΔE 2 due to the similar charge on cathode and anode (i.e., ω α % ω β ). Therefore, the voltage window of SSCs depends on the dissociation energy of electrolyte. [22] In aqueous SCs, the different mechanisms associate with each other, collectively affecting their overall performance. According to Equation (1), work function and dynamic mechanism of HER/ OER can simultaneously decide the voltage window. In recent years, some studies have also reported other mechanisms that can improve the SCs voltage window, but the rationality and the relation with accepted mechanism are supposed to be further discussed.

Strategies for Constructing Aqueous
Supercapacitors Beyond 2.0 V

Modifying Electrode
Based on the foregoing discussed theoretical mechanisms, the novel strategies of electrode modification are widely studied to replace the conventional methods (e.g., electrodes mass balance [24] and electrode surface passivation [2b] ), which effectively push aqueous SCs voltage window over 2.0 V. These strategies can be classified into structure engineering, metal cations doping, and constructing advanced composites.

Structural Engineering
The morphology of electrode materials directly affect the overall electrochemical performance of SCs, especially the specific capacitance. Consequently, many rational designs of electrode structure were attempted to expand the voltage window in aqueous SCs.
MnO 2 is an appealing cathode material in SCs due to the high theoretical capacitance (1371 F g À1 ) and the immanent high oxygen evolution potential. The initial progress of aqueous SCs with >2.0 V voltage window was used on MnO 2 -based electrode. Yu and co-workers [25] assembled SSC with honeycomb porous MnO 2 synthetized by solution reaction, which delivered an voltage window of 2.2 V in neutral electrolyte. The honeycomb structure was composed of interconnected MnO 2 nanosheets with thickness of 3-7 nm grown vertically on the carbon nanofibers, forming honeycomb-like nanopores of 20-50 nm, which was exhibited by the scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images ( Figure 3a). This porous-open-structure of honeycomb-like electrode provided a vast number of electrochemical active sites for Mn 3þ /Mn 4þ redox reaction, which required more transferred electrons involved in the redox reactions, resulting in a potential increasing. However, a severe shortcoming was the capacity retention can only reach 76% after 3500 cycles. The similar morphology MnO 2 -based aqueous SCs were further investigated by other researchers. Wang et al. [26] and Dai and co-workers [28] both fabricated porous MnO 2 -based electrodes by the facile electrochemical deposition method (Figure 3b). The corresponding ASCs achieved the voltage window of 2.2 and 2.4 V with improved capacity retention of 88.6% and 88.7% after 5000 cycles, Figure 2. a) Relationship between WFD and voltage window of ASCs. b) Schematic presentation of ASCs using CKMO as positive electrode and metal oxide as negative electrode and contribution of their WFD to voltage window. a,b) Reproduced with permission. [21] Copyright 2020, Elsevier. c) Schematic of ASC based on graphene/MnO 2 cathode and graphene/MoO 3 anode. d) Work function of different metal oxides. e) Relationship between voltage window and WFD of MnO 2 cathode and MoO 3 anode. c-e) Reproduced with permission. [22] Copyright 2013, Wiley-VCH. respectively (Figure 3c-e). The improved cycling retentions were obtained by these reasons: 1) electrochemical deposition as the chosen synthetic method was able to provide stronger electrodecollector bonding force than those prepared by the solution reaction; 2) the low-crystallinity honeycomb MnO 2 had less content of structure water molecules than those in other porous MnO 2 with large interlayer tunnel space. The basic metal cation can hop between H 2 O and OH À sites of structure water molecules inserted into the MnO 2 interlayer. In other words, structure water promoted the electrolyte cations migration on the surface of MnO 2 electrode, which can undertake faster electron transfer at high potential, e.g., over 2.0 V. Thus, MnO 2 electrode with high content of structure water can effectively avoid structural collapse during the long-term high-voltage window cycling.
Inspired by the porous-open-structure of these MnO 2 -based electrode, the aqueous SCs (>2.0 V) fabricated by other materials electrodes with the approximate structure have also been studied. For example, Tripathi and co-workers [29] utilized hydrothermal method to synthesize porous flower-like NiCo 2 O 4 quantum dots (NCO-QDs) electrode. The aqueous ASCs with prepared NCO-QDs as cathode and reduced graphene oxide as anode showed 2.5 V voltage window and a good energy density of 69.5 Wh kg À1 along with excellent power density of 2.22 kW kg À1 , exhibiting 86% capacitance retention after 1000 cycles. In addition to the unique porous structure, the QDs can also provide huge number of active sites for redox reaction because of their superior conductivity, small size, and large specific surface. In addition, Wang et al. [27] created a porous boron-doped diamond (BDD) film by a self-assembly seeding approach via microwave plasma chemical vapor deposition (MPCVD) technique. The fabricated aqueous SSCs had a wide voltage window of 2.6 V and a great cycling stability of 91.5% retention of initial capacity after 10 000 cycles (Figure 3f,g). However, sp 2 -based carbon electrodes SSCs had a relatively poor power density of 5.46 Wh kg À1 with 2.9 kW kg À1 .
Although the aqueous SCs with structure engineering acquire >2.0 V voltage windows benefited from massive redox active sites provided by the porous-open-structure electrodes, their stabilities of capacitance retention were mostly unsatisfied. The increased active sites not only raised the electrochemical activity but also may promote the adsorption of H þ /OH À from electrolyte onto electrode surface, which increased HER/OER activity either. Therefore, the structural collapse caused by the gas Figure 3. a) SEM and TEM images of honeycomb porous MnO 2. Reproduced with permission. [25] Copyright 2014, Elsevier. b) Schematic of the main procedures for fabricating porous δ-MnO 2 . c) CV curves at 40 mV s À1 in varying voltage windows. d) Cycling test at 4 A g À1 and the corresponding Coulombic efficiency. e) GCD curves at varying current densities. b-e) Reproduced with permission. [26] Copyright 2020, American Chemical Society. f ) Schematic of SSC structure and CV test of planar BDD electrode and porous BBD electrode. g) Stability test of SSC device. f,g) Reproduced with permission. [27] Copyright 2020, Elsevier.
release of water decomposition can reduce the capacity performance during the long-term charging-discharging process. Overall, the single structural engineering is a directly effective way to enhance the voltage window; meanwhile, it still needs to overcome the hurdle of capacitance sacrificing.

Metal Cations Doping
To enlarge the voltage window without capacitance sacrificing, metal cations doping on electrodes through electrochemical technique has been proved as a feasible method by many researchers, which mainly depends on decreasing the activity of OER and HER.
In the aspect of positive electrodes, all studies have been focused on manganese oxides-based materials. As the typical manganese oxides-based cathode materials, metal cations doped MnO 2 has been extensively studied. MnO 2 exhibits multiple crystal phases (e.g., α-, β-, and δphase), and their OER activity strongly depend on the crystallographic structures, which is following an order of α-MnO 2 > β-MnO 2 > δ-MnO 2 . [30] Prabakar and co-workers [31] utilized a facile hydrothermal technique to incorporate K þ into cryptomelane α-MnO 2 . The prepared α-KMnO 2 had an enhanced stabilization; however, the potential window of α-KMnO 2 cathode was still limited within 1 V due to the high OER activity. To achieve higher potential window, more studies of metal cations doped cathode focused on the birnessite δ-MnO 2 . In this case, metal cations doping mainly refers to metal cations insert into manganese oxides framework via cyclic voltammetry (CV) technique at a large potential range in the corresponding cation-containing electrolytes, which is called phasetransformation electrochemical activation or electrochemical oxidation. For example, Xia and co-workers [32] utilized electrochemical oxidation to synthetize birnessite Na 0.5 MnO 2 electrode that was converted from spinel Mn 3 O 4 . In addition, approximate Na 0.25 MnO 2 electrode were fabricated in Xue and co-workers [15a] via similar electrochemical oxidation procedure with the electrolyte at a relatively lower concentration. The corresponding aqueous SCs (Na 0.5 MnO 2 //Fe 3 O 4 @C and Na 0.25 MnO 2 //Na-C) were determined to deliver ultrahigh voltage window of 2.6 and 2.7 V, respectively, in which the Na x MnO 2 cathode exhibited the same potential window (0-1.3 V). The Na x MnO 2 was formed during the CV scan process. In charging stages, the Mn 2þ in tetrahedral Mn 3 O 4 was dissolved, whereas the unstable Mn 3þ was oxidized to Mn 4þ . In the meantime, water molecules inserted into the interlayer space. In the discharge stages, the Na þ intercalated into the internal tunnel space resulting in partial Mn 4þ species being reduced to Mn 3þ ( Figure 4a). The Na þ introduced into the Na x MnO 2 structure was detected via the X-ray photoelectron spectroscopy (XPS) spectrum ( Figure 4b). After several CV cycles, the final Na x MnO 2 with octahedral structure units was formed and the corresponding elements were observed by energy dispersive X-ray (EDX) spectrum ( Figure 4c). The charging-discharging mechanism for MnO 2 -based cathode of metal cations insertion can be presented by the following equation where C represents alkali metal cations and x refers to amount of transferred electron; the content of inserted metal cation (i.e., the value of x) can be adjusted via changing the parameters of three electrode system under the CV technique, such as scan rate (mV S À1 ), cycle potential range, CV numbers, and electrolyte concentration.
The aforementioned strategies not only aroused the development for other positive electrode materials of Na x MnO 2 -based [34] but also pushed forward the electrode studies of manganese oxide-based with different metal cations intercalation (i.e., C x Mn y O z ), which actually aided the relevant aqueous SCs in achieving ultrahigh voltage window (mainly >2.4 V). Liu and co-workers [3a] synthesized Ni 0.25 Mn 0.75 O solid-solution electrode via a facile hydrothermal method followed by annealing, which had the prominent stability due to the synergistic effect between Mn and Ni ion through oxygen bonding bridge. The facile formation of Ni 0.25 Mn 0.75 O solid-solution structure was due to the similar cubic crystal structure of NiO to MnO and the very close ionic radius of Ni 2þ (0.069 nm) and Mn 2þ (0.067 nm). Afterward, through a CV phase-transformation electrochemical activation, the pristine Ni 0.25 Mn 0.75 O was converted to LiNi 0.5 Mn 0.15 O 4 of low crystalline. The CV activation was an OER gradually disappearing process, in which OER existed at %0.8 V in the first cycle and eventually vanished at the seven cycle, finally resulting in a stable wide electrochemical window of 0-1.4 V ( Figure 4d). Excepting Na þ and Li þ , K þ -intercalated MnO 2 had also been reported. [21,33] For instance, the aqueous SCs of K 0.6 MnO 2 // K-C fabricated by Xue and co-workers delivered 52.8 Wh kg À1 at 0.5 kW kg À1 in a stable voltage window of 2.4 V (Figure 4e,f ). [33a] Although the ionic size of K þ is larger than those of Na þ and Li þ , the radius of hydrated K þ (3.31 Å) is slightly smaller than those of Na þ (3.58 Å) and Li þ (3.82 Å) due to the relative weak K þ -H 2 O interaction. [35] As a consequence, hydrated K þ exhibits higher mobility, which may lead to a higher conductivity in this type of intercalation electrode. Unfortunately, there still lack the specific comparison of intercalation cathodes of different metal cations under the same preparation process. These large cathodic potentials were attributed to the enhanced onset overpotential for OER, as the insertion of metal cations into the manganese oxide-based is an energy requiring process, which restricts the OER activity by competing against oxygen evolution process. In addition, the work function of cathode materials decreased when metal cations intercalate into MnO 2 structure due to the the incremental content of Mn 3þ , that is, the lower average oxidation state of Mn. The work function of MnO 2 (4.4 eV) are lower than other transition metal oxides, thus, lower work function of MnO 2 -based cathode leads to an expansion of WFD that enlarges the entire voltage window. [32,34] As for negative electrode, the cations doping was basically focused on carbon materials. [15a,33a,34a,36] Thus, this type of cations doping can be interpreted as the cation adsorption on carbon-based anode via an electroreduction method of CV cycle in a low potential range. On the surface of carbon materials, there are many defective sites existing in atomic structure of sp 2 -C, sp 3 -C, and sp 3 -O. Metal cations can adsorb on the surface carbon-based electrode to replace these defective sites by a thermodynamically favorable process, which is concluded by the density function theory (DFT) simulation of Gibbs free energy of adsorption (i.e., ΔG ads < 0) ( Figure 5a). The adsorbed metal cations acted as a physical barrier to block adsorption of H þ , which effectively decreased the HER activity. In addition, part of the O-containing groups that may provide active sites for HER can be removed during the electroreduction cycles, such as C═O, O─H, C─O─C. Xue et al. prepared the electroreduction Na─O─C porous carbon as aqueous SCs anode with the potential window of À1.4 to 0 V. During the electrochemical reduced process, the amount of C═O was decreased. Thus, Na þ was hard to remove, and confirmed to exist on the anodic surface via EDS and XPS spectrum (Figure 5b,c). In general, large HER overpotential for carbon-based anode can be realized by the joint efforts of metal cations adsorption and reduced oxygen functionality in aqueous electrolyte (Figure 5d). Another advantage of metal cation incorporation was enhancing surface wettability of the carbon-based anode to increase the ion-migration rate, which favors the effective formation of electric double layer. The high-wettability was affirmed by the dynamic contact-angle measurements (Figure 5e). In addition to the studies of carbonbased anode, Song et al. [37] used the electro-deposited molybdenophosphate material as negative electrode, which delivered the ultralow cutoff potential of À1.5 V (Figure 5f,g). Interestingly, the electro-activation prepared K 1.55 Mo 2 O 4.2 PO 4 built by the corner-sharing MoO 6 octahedra and PO 4 tetrahedra had the similar internal open tunnels structure with that of C x MnO 2 . The insertion of K þ was not only an energy-required process that increased the HER overpotential but also introduced the oxygen defect into the lattice. Hence, the HER overpotential was further enlarged by the larger resistance polarization and bandgap value caused by the introduced oxygen defect. [1d] According to the aforementioned discussion, metal cations doping can be mainly divided to cathodic metal cations intercalation and anodic metal cations adsorption. For cathodic aspect, unfortunately, the correlation between the content of inserted metal cation and cathodic potential window has not been systematically studied. In general, the specific capacitance can be promoted with high content cations by increasing active sites of pseudocapacity. Nevertheless, the excessively high content of cations may destroy the reversibility of electrode redox reaction, which decreases the cycling stability in a wide operating window. Furthermore, the comparisons of electrochemical performance of the electrode with  various intercalated/adsorbed metal cations are worthy of further study. As for anodic aspect, the study of metal cation adsorption faced the similar issues of cathodic ones. Meanwhile, the selection of both positive and negative electrode materials for metal cations doping are very limited. More types of materials can be attempted with this effective method; for instance, the molybdenophosphatebased materials provided us a new sight to search more types of anode with large anodic potential window.

Constructing Advanced Composites
The strategies of electrode modification that generate homogenous monocomponent materials may suffer from poor cycling stability (e.g., those by structure engineering) or very limited material options (e.g., those by metal cations doping). Therefore, the rational construction of composite electrode has been attempted in aqueous SCs. The synergistic effect of rationally designed composites in aqueous SCs can offset the weakness of single substance by taking advantage of the superior properties from different components to achieve the ultrahigh voltage window. For example, Yang and co-workers [38] conducted a composite positive electrode, denoted as CSN-PB/MnO 2 , by an in situ electrochemical growth of Prussian blue on the carbon cloth, which was followed by a electrodeposition of MnO 2 (Figure 6a). Prussian blue is a typical coordination supramolecular network with the cube lattice of layered and interconnected structure. The as-prepared aqueous SC was able to work in a stable voltage window of 2.4 V, wherein the CSN-PB/MnO 2 cathode reached a high potential window of 1.4 V due to the synergistic effect of Prussian blue and MnO 2 in two aspects. First, during the process of electrochemical in situ growth of Prussian blue, the metal cations were inserted into the specific cube lattice. In the chargingdischarging process, the preinserted metal cations intercalated into MnO 2 layer benefiting from the large transfer tunnel and the intimate contact between two constituent components. Thus, the OER overpotential was increased by the extra metal cations intercalation, which shared the alike mechanism with that in metal cations doping strategy of MnO 2 -based cathode.  Second, the open framework structure of CSN-PB promoted the specific surface area and roughness of MnO 2 . Wherein, the large specific surface area can provide considerable active sites to enhance the potential window, which was similar to the aforementioned strategy of structure engineering. Moreover, the rougher MnO 2 with CSN-PB framework supporting can keep electrode structure away from pulverization during the longterm charging-discharging cycling. Subsequently, the CSN-PB/ MnO 2 //AC device exhibited a good capacitance retention within the voltage window of 2.4 V after 20 000 electrochemical cycles at the current density of 20 A g À1 . In addition to the synergistic effect of structure engineering and metal cations doping, other synergistic mechanisms were also explored to aid the aqueous SC in achieving the voltage window over 2.0 V. Fei and co-workers [39] utilized plasma enhanced chemical vapor deposition (PECVD) technique to fabricate graphene quantum dot (GQD)/MnO 2 heterostructural cathode by GQD in situ growing on the surface of MnO 2 nanosheet, in which the sizes of GQD were were % 2-3 nm (Figure 6b). The heterostructure was established through Mn─O─C covalent bonds with PECVD technique, which was inexistent with a conventional hydrothermal method. In the heterostucture, a built-in electric field was formed due to the WFD between GQDs (%5.2 eV) and MnO 2 (4.4 eV) (Figure 6c). The built-in field was incapable to be completely screened because of the ultrasmall sizes, resulting in the extension of potential window. Meanwhile, the Mn─O─C covalent bonds maintained the structural stability during the long-term electrochemical cycling. Enlightened by this case, we deem more heterostructural components electrodes are worth exploiting in aqueous SCs domain. In addition, the composite electrodes without synergistic effect had been used in >2.0 V aqueous SCs. Qin et al. [40] presented a new thought of constructing the composite electrode by electrodepositing the alkali-type double-salt composites (Zn/Zn 4 SO 4 (OH) 6 ·4H 2 O) on the porous carbon cloth with basic oxygen-containing group (OCC). During the charging process, the electrolyzation of water generated the H þ and OH À simultaneously, in which hydrogen can store in the porous carbon via the inhibition of basic functional groups and the OH À were anchored on the carbon cloth by the formation of Zn/Zn 4 SO 4 (OH) 6 ·4H 2 O precipitation (Figure 6d). However, this alkali-type double salt only precipitated under the condition of pH higher than 5.3. During the discharging process, the  hydrogen stored in the porous carbon was oxidized into H þ , decreasing the electrolyte pH value near the electrode. Thus, the precipitation would dissolve when pH value was below 5.3, which conducted a reversible precipitation/dissolution process. This Zn/Zn 4 SO 4 (OH) 6 ·4H 2 O@ OCC anode expended the potential window to À1.7 V through effectively depressing HER activity. In addition to the double-salt composite electrode, a carbon-based composite electrode with metallacarboranes as sacrificing mediators was fabricated by Ruiz-Rosas et al. [41] They utilized a conventional method to physically connect metallacarboranes (e.g., Na[Co(C 2 B 9 Cl 2 H 9 ) 2 ]) into carbon electrode with polytetrafluoroethylene (PTFE) as the binder and carbon black as conductivity promoter. The metallacarboranes exhibited the tuned redox potential closed to HER and OER potential of the electrolyte, which acted as electron-consuming agent to avoid the water decomposition (Figure 6e). Unfortunately, the dissolved anion of metallacarboranes during the reversible redox reaction was hard to strongly absorb onto the carbon electrode due to the lack of any chemical bonding. Indeed, the reversible precipitation of alkali-type double salt and the mediator-sacrificing composite electrodes brought us the new creative ideas for constructing aqueous SCs of voltage window upon 2.0 V, but their system electrochemical reversibility showed clear flaws. Similarly, most of the iron oxides-based electrodes with broad potential window suffer from poor cycling stability. Xia and co-workers. [32] synthesized the carbon-coated Fe 3 O 4 anode (i.e., Fe 3 O 4 @C composite electrode) via a facile hydrothermal method followed by annealing, which delivered a wide potential window of 1.3 V along with the excellent cycling stability of 92% capacitance retention after 10 000 CV cycles. The ultrathin carbon layer worked as a protection layer that effectively suppressed the dissolution of iron into the electrolyte. Accordingly, this carbon-coated composite electrode provides an approach to improve the stability of iron oxides-based electrode by constructing metal oxides composite electrode with ultrathin carbon protection layer.
With the rational design, the advanced composite electrodes can combine the advantages of different strategies of electrode modification, or exhibit the innovative mechanism (e.g., builtin electric field), delivering the high potential in aqueous SCs. Moreover, the synergistic effect of components can defend the electrode structure against collapsing or dissolution during long-playing electrochemical cycling. Howbeit, the complicated preparation procedure and the high-cost experimental technique (e.g., PECVD) extraordinarily slow down their developing steps. Meanwhile, the undesired electrochemical reversibility of other composites electrodes without synergistic effect is another unsolved aporia.
By summarizing the aforementioned strategies of electrode modification and their electrochemical performances of the corresponding aqueous SCs (Table 1), we propose our views as listed: 1) the fabrication of porous-open-structure electrode can directly increase the potential window, which, however, suffers from the undesirable electrochemical cycling stability; 2) the doping of metal cation into electrode is an effective way to www.advancedsciencenews.com www.small-structures.com enlarge voltage window as well as to improve cycling stability, which have been extensively reported. But the selectable materials are still very limited; 3) constructing composites electrodes with synergistic effect brings us new inspirations to achieve ultrahigh voltage window in aqueous SCs. However, the high complexity and technicality of the fabrication techniques will be severe challenge for further investigation. In addition, improving the electrochemical reversibility of composites electrodes without synergistic effect will be another tough challenge simultaneously.

Optimizing Electrolyte
At present, commercial SCs exhibit a stable voltage window (>2.5 V) using organic electrolyte. For further large-scale applications, the optimized aqueous electrolyte along with natural merits of safety, environmental friendliness, and low cost have been extensively studied to replace the organic electrolyte.

"Water in Salt" Electrolyte
The properties of SCs are determined by the cooperation between electrode and aqueous electrolyte. Although the aqueous SCs with aforementioned electrode-modification strategies had the competitive voltage windows, their complicated preparation conditions increase the difficulties for practical scale-up production. Accordingly, the investigations of novel aqueous electrolytes were concentrated on assembling aqueous SCs with high voltage window comparable to organic SCs, in which the electrodes can be prepared with facile procedure, e.g., pure metal oxides/ nitrides and activated carbon. "Water in Salt" electrolyte denotes the super-concentrated aqueous electrolyte, in which salt exceeds water solvent in the content of weight or volume (i.e., salt/water ratio > 1). The WIS electrolyte was first reported by Wang and co-workers. [42] in the LIB with 2.3 V voltage window. They chose lithium bis (trifluoromethane sulfonyl) imide (LiTFSI) to be dissolved into water with ultrahigh concentrations (> 20 molality) because of its high solubility and hydrolysis stability. In the WIS electrolytes, the extremely high salt concentration leads to an anioncontaining cation solvation sheath with low water molecules concentration (Figure 7a). Almost all the water molecules of electrolyte were locked down in the cation solvation sheath by strong coordination with metal cations via the Lewis-basic oxygen atoms, which intensively suppressed electrochemical activity of water molecules, resulting in a stable and high voltage window. Inspired by this case, ultrahigh concentration LiTFSI as the typical WIS electrolyte were initially used in aqueous SCs to achieve the voltage window of > 2.0 V. When the concentration of LiTFSI runs up to 5 molality, the binary electrolyte can be defined as WIS electrolyte due to the salt/water ratio of beyond 1. Nakanishi and the co-workers [45] chose 6 M KOH, 1 M LiSO 4 , 2.5 M LiTFSI (3.9 mol kg À1 ), and 5 M LiTFSI (18.7 mol kg À1 ) as electrolytes to conduct the AC//AC SSC. Wherein the AC electrode showed the stable potential window of 1.0, 1.8, 2.0, and 2.5 V with 6 M KOH, 1 M LiSO 4 , 2.5 M LiTFSI, and 5 M LiTFSI electrolytes, respectively. For 5 M LiTFSI electrolyte system, except the reduced water activity caused by the solvation sheath of low water molecules, a LiF-rich passivation layer on the electrode surface formed by the LiTFSI reductive decomposition blocked the H 2 evolution to further enlarge the HER overpotential. Meanwhile, the capacitances of electrode in 2.5 M LiTFSI electrolyte and 5 M LiTFSI electrolyte showed approximate value at 5 mV s À1 and were slightly lower than that of in 1 M LiSO 4 electrolyte due to the large-size TFSI anion, which proved the solid decomposition Figure 7. a) Illustration of the evolution of the Li þ primary solvation sheath in diluted and WIS solutions. Reproduced with permission. [42] Copyright 2015, American Association for the Advancement of Science. b) Potential window and specific capacitance of MnO 2 electrode in SIW and WIS electrolytes. Reproduced with permission. [43] Copyright 2016, Elsevier. c) Snapshots of 21 m WIS electrolyte and 5 m AWIS electrolyte. d) Conductivity and viscosity of AWIS electrolytes with different concentrations. c,d) Reproduced with permission. [6] Copyright 2018, Royal Society of Chemistry. e) Ternary phase diagram of LiTFSI/H 2 O/ACN hybrid electrolytes at room temperature. Reproduced with permission. [44] Copyright 2019, Wiley-VCH. f ) ESWs of different concentrated NaClO 4 electrolytes on stainless-steel electrodes at a scan rate of 10 mV s À1 . g) Price comparison of various salts and solvents, as well as the corresponding electrolytes. f,g) Reproduced with permission. [11b] Copyright 2019, Royal Society of Chemistry.
www.advancedsciencenews.com www.small-structures.com product of LiTFSI exerted slight effect on capacitance performance. However, the 5 M LiTFSI electrolyte exhibited poor ionic conductivity (%13 mS cm À1 ) due to the low free-water molecule content, compared with 1 M LiSO 4 (%60 mS cm À1 ) and 2.5 M LiTFSI (>50 mS cm À1 ). In addition to the electrochemical characterization contrast of carbon-based electrode between 5 M LiTFSI electrolyte and conventional "Salt in Water" (SIW) electrolyte, [45,46] the MnO 2 -based electrode in 5 M LiTFSI electrolyte also delivered better potential window and capacitance performance than that of in some conventional SIW aqueous electrolytes (Figure 7b). [43] In addition, many higher concentration LiTFSI electrolytes (>5 M) were used in aqueous SCs of >2.0 V, [11a,47] in which the saturation concentration of LiTFSI is capped at %21 M. For instance, the MnO 2 //VNS SCs in 21 M LiTFSI electrolytes even delivered a voltage window of 3.0 V with the large energy density of 61.5 Wh kg À1 .
[47c] However, the low ionic conductivity and high viscosity of super-concentrated LiTFSI WIS electrolytes led to a poor rate performance, which was caused by the strong electrostatic attraction between cations and anions. Hence, Yan and co-workers [6] blended acetonitrile (ACN) with LiTFSI WIS electrolyte to form acetonitrile/WIS (AWIS) hybrid electrolyte, which weakened the cation-anion electrostatic attractions benefiting from the spatial isolation by ACN molecules, resulting in the improvements of ionic conductivity and viscosity (Figure 7c,d). Meanwhile, the strong coordination between metal cations and water molecules were still preserved. Although ACN is a highly flammable solvent, it was interesting to find out the AWIS electrolytes were nonflammable when the concentration was ≥5 M. Subsequently, Yan and co-workers [44] further summarized the relationship between the molar fraction ratio of AWIS hybrid electrolyte and its compositive properties (e.g., solubility, flammability, and conductivity) (Figure 7e). These series of work indicated AWIS electrolytes can display better ionic conductivity and viscosity performance than typical LiTFSI WIS electrolytes, and their safety were also controllable. Despite the WIS/AWIS electrolytes of LiTFSI had shown their advantages on constructing aqueous SCs of ultrahigh voltage window, the high cost of LiTFSI as well as potential safety issue may limit their future widespread applications. Accordingly, other highly soluble salts with cheaper prices and inherent higher ionic conductivity had been studied as new type WIS electrolytes of aqueous SCs. Yan and co-workers [11b] fabricated the carbon-based aqueous SSCs with 17 M NaClO 4 -based WIS electrolyte, which delivered 2.3 V voltage window along with outstanding cycling stability. The linear sweep voltammetry (LSV) measurement of stainless steel two electrodes in 17 M NaClO 4 electrolyte showed a large ESW of 2.8 V, which was 0.5 V higher than that of 2 M NaClO 4 (Figure 7f ). Furthermore, in contract with 21 M LiTFSI, 17 M NaClO 4 exhibited lower viscosity (5.0 vs 30.2 mm 2 s À1 ) and higher ionic conductivity (64.2 vs 8.2 mS cm À1 ). More importantly, the price of NaClO 4 (0.49 $ g À1 ) is much lower than that of LiTFSI (7.18 $ g À1 ). Correspondingly, the price of 17 M NaClO 4 electrolyte (0.33 $ g À1 ) is not only cheaper than 21 M LiTFSI electrolyte (6.16 $ g À1 ) but also lower than those of the current commercial organic electrolytes (Figure 7g). Except 17 M NaClO 4 electrolytes, [11b,48] there were other WIS electrolytes with similar cheap price aiding the corresponding aqueous SCs achieving the voltage window over 2.0 V, such as 12 M NaNO 3 [49] and 8.96 M LiCl WIS electrolyte. [50] Unfortunately, there were a significant gap in voltage window between the aqueous SCs with these WIS electrolytes and those with typical LiTFSI WIS eletrolytes. Accordingly, by combining 8 M NaClO 4 with ACN solvent, the voltage window of the aqueous SCs with NaClO 4 /(H 2 O) 1.5 /(ACN) 2.4 WIS electrolyte reaching 2.5 V was reported by Yan and co-workers. [51] In general, WIS electrolyte as a novel electrolyte can provide an ultrahigh voltage window comparable to commercial organic electrolyte even when utilizing conventional materials electrode. The typical WIS electrolytes (LiTFSI) suffer from their low ionic conductivity and high viscosity, which can be ameliorated by introducing a small quantity of organic solution (ACN). However, the AWIS hybrid electrolytes require strict moisturefree encapsulation to maintain the optimized water/organic solvent ratio to avoid potential safety risk, which increase the cost and complexity of practical production. Meanwhile, the studies of other highly soluble salts with cheaper price are evoked due to the high dosage of super-concentrated salt. Unfortunately, the voltage window performances of these WIS electrolytes are not satisfied as those of LiTFSI WIS/AWIS electrolytes. Thus, the further research direction can focus on searching for the WIS electrolytes with low cost as well as higher voltage window and perfectible electrochemical performance.

Novel Mixed Electrolyte
In addition to the ultrahigh concentration WIS electrolyte, some aqueous SCs with mixed low-concentration electrolyte also delivered > 2.0 V voltage window. These mixed electrolytes are composed of multi-electrolytes separated by functional membrane or the electrolyte with additive redox species.
Compared with neutral electrolytes, carbon-based electrodes exhibit higher HER/OER overpotential in alkaline/acidic electrolyte, respectively, due to the low concentration of H þ in alkaline electrolyte and OH À in acidic electrolyte. Moreover, the capacitive performance of carbon-based electrode is also limited in neutral electrolytes due to the low ionic conductivity, the electrochemical adsorption of hydrated anions with large radium, and negligible pseudocapacitive contribution under neutral condition. [52] In this case, Wu and co-workers [53] prepared LiMn 2 O 4 //NBC aqueous SCs (2.3 V) with the mixed electrolyte separated by a modified Nafion membrane, that is, a K þ conductive membrane. The mixed electrolyte was composed with 1 M Li 2 SO 4 /0.3 M K 2 SO 4 on cathodic side and 2 M KOH/0.3 M K 2 SO 4 on anodic side (Figure 8a). In contrast, if only single Li 2 SO 4 was used, the aqueous SCs presented lower energy density; and if only KOH electrolyte was used, the cathode presented poor cycling stability. Thus, it is clear the special mixed electrolyte aided the aqueous SCs in exhibiting exceptional energy density as well as excellent cycling performance due to the full use of anodic capacitance and low potential without sacrificing cathodic electrochemical stability (Figure 8b,c). Despite the cation conductive membrane realized the configuration of mixed multielectrolytes, the expensive price of Nafion membrane and the complicated converting process of its conductive cation. Considering the issues of cation conductive membrane, Zhang and co-workers [54] proposed a compact Janus membrane with bipolar to replace the cation conductive membrane in the similar www.advancedsciencenews.com www.small-structures.com SCs structural configuration. A carbon-based aqueous SC was fabricated, in which the constituents of the mixed electrolyte (1 M H 2 SO 4 /2 M KOH) were separated by the bipolar Janus membrane. The Janus membrane was composed with sulfonated polystyrene (on H 2 SO 4 side) and quaternized polystyrene (on KOH side). The K þ cations and SO 4 2À anions were blocked by the bipolar membrane, whereas H þ /OH À still transfer free (Figure 8d). Benefiting from the mixed electrolyte system, a successive voltage window of 2.2 V was obtained. Unfortunately, the as-prepared aqueous SCs had a poor cycling stability (62% capacitance retention after 2200 cycles), which was likely caused by the degradation or substitution of the functional groups on the used Janus membrane. In addition, another mixed electrolyte was presented by Xia and co-workers, [55] which was formed by pre-adding pseudocapacitance redox species electrolyte into main electrolyte before electrochemical operation. They used Co 3 O 4 anode and carbonbased cathode to fabricate ACC@RGO//Co 3 O 4 aqueous SC of 2.2 V with 2 M Li 2 SO 4 /0.1 M CoSO 4 mixed electrolyte. The anode Figure 8. a) Structural illustration and energy storage mechanism of the as-fabricated NBC//LiMn 2 O 4 SSC with the alkaline-neutral electrolyte. b) Individual CV curves of NBC and LiMn 2 O 4 electrodes in 2 M KOH and 1 M Li 2 SO 4 aqueous electrolyte solutions, respectively. c) CV curves of the SSC device with the alkaline-neutral electrolyte at different scan rates. a-c) Reproduced with permission. [53] Copyright 2019, Royal Society of Chemistry. d) Schematic of the ASC device using Janus membrane and asymmetric electrolytes. Reproduced with permission. [54] Copyright 2019, Elsevier. e) ESWs collected in 2 M KOH and 2 M Li 2 SO 4 -0.1 M CoSO 4 electrolyte solutions by LSV tests using Pt foil as current collectors. Reproduced with permission. [55] Copyright 2020, American Chemical Society. pseudocapacitance reaction mechanism was confirmed (i.e., Co 3 O 4 þ 2H þ þ 2e À ↔ 3CoO þ H 2 O). The pre-adding Co 2þ from 0.1 M CoSO 4 not only kept the balance of Co 2þ concentration variation to stabilize the anode electrode but also enhanced H þ desorption energy barrier, which inhibited water decomposition and effectively increased the voltage window ( Figure 8e). 0.1 M was considered as the optimized concentration of CoSO 4 additive, because the potential window would not further increase with higher concentration. However, the optimized additive concentration existed uncertainty due to the lack of systematic electrochemical performance comparison with different additive concentration. Furthermore, seeking and maintaining the perfectible additive concentration would be harder for for large-scale production. The mixed electrolyte separated by functional membrane show their potential in high voltage window aqueous SCs. The current problems are caused by the immature membrane techniques. Wherein, Janus membranes suffer from the unreliable functional group and cation conductive membranes undergo the complex preparation process and the expensive market price. For the additive-type mixed electrolyte, it is hard to maintain their electrochemical stability due to the constantly changing additive concentration during the electrochemical cycling process.
Currently, these optimized novel aqueous electrolytes can be considered as promising candidates to replace current commercial organic electrolyte. The electrochemical performances of > 2.0 V aqueous SCs with aforementioned optimized electrolytes are shown in Table 2. However, the existing drawbacks still need to be further solved as follows: 1) WIS electrolyte should explore more suitable salts in cheaper price or better strategies to solve their low conductivity and high viscosity without sacrificing safety index. 2) The functional membranes for separating mixed electrolyte need more stable functionality along with the cheaper price and fabrication process. Meanwhile, new approach can be involved to keep the concentration stability of additive-type mixed electrolyte.

Conclusion and Outlook
In this review, we initially introduced the theoretical mechanism in regard to HER/OER and work function as well as their effects on voltage window of SCs. Afterward, the achievements of aqueous SCs with >2.0 V voltage window were thoroughly summarized. The progress and the unsolved issues of these achievements were analyzed in the aspect of different strategies, including constructing structural engineering, metal cations doping, and advanced composites electrode; preparing WIS and novel mixed electrolyte. Despite these novel advanced strategies brought the ultrahigh voltage window (>2.0 V) for aqueous SCs, some flaws of these strategies impeded aqueous SCs replacing the organic SCs in the commercial market. To sum up, the current achievements, existing challenges, and future development directions of aqueous SCs with ultrahigh voltage window are pointed out as follows: 1) The mechanism studies on voltage window of aqueous SCs were basically located on water decomposition potential. For ASCs device, work function is another important reference to adopt electrode materials. In addition, other field mechanisms are worth of more exploration, such as built-in electric field of heterostucture. [37] 2) Fabricating open-porous-structure electrodes is facile and straight way to raise the voltage window but suffer from the poor electrochemical cycling performance. The enhancement of voltage window is in virtue of the increased active sites. However, the increased electrochemical reaction intensity synchronously promotes easy degree of electrode structural collapse. Hence, based on this structural framework, constructing composites electrode with synergistic supportive effect should receive more attention. In addition, benefiting from the rational cooperation of components, some advanced composites electrodes with multi-synergistic effect can provide a stable high potential window. The simplification of technique and procedure of electrode preparation are the key factors for their further development.
3) The metal cations doping on both positive and negative electrode materials can effectively enlarge the voltage window without sacrificing capacitance retention. However, the specific effects of metal cation doping content on voltage window and other electrochemical performance have been rarely studied. Meanwhile, the influence of different species of metal cation on voltage window ought to receive more discussion. In addition, the used electrode materials of this strategy are basically concentrated on manganese oxide-based cathode and carbonbased anode, which may decrease the applied domain. Thus, more types of positive and negative electrode materials are supposed to be investigated with the strategy to meet future multifield applications. 4) The super-concentrated WIS electrolytes and the mixed multi-electrolytes separated by functional membrane both showed huge potential in replacing the organic electrolyte of current commercial SCs. The aqueous SCs assembled by these novel electrolytes and conventional activated carbon electrodes delivered comparable high voltage window of commercial organic SCs. Nonetheless, the widely used WIS electrolytes of LiTFSI suffer from the higher cost than that of the commercial organic electrolytes. Therefore, the future research direction of WIS electrolytes can focus on exploiting the appropriate WIS electrolytes with cheaper price, meanwhile, maintaining superiority of voltage window. As for separated mixed electrolyte system, the unreliable lifespan of functional membranes is urgent to be ameliorated, as well as their complex preparation process and high cost. To date, the strategies of electrode modification and electrolyte optimization in >2.0 V aqueous SCs were typically used separately. We believe the combination of the advanced strategies about electrode and electrolyte materials would become the new study trend of aqueous SCs with ultrahigh voltage window. We hope this review article could provide beneficial inspiration for constructing great aqueous SCs with higher voltage window and better electrochemical performance.