Metal chalcogenides for potassium storage

Potassium-based energy storage technologies, especially potassium ion batteries (PIBs), have received great interest over the past decade. A pivotal challenge facing high-performance PIBs is to identify advanced electrode materials that can store the large-radius K + ions, as well as to tailor the various thermodynamic parameters. Metal chalcogenides are one of the most promising anode materials, having a high theoretical specific capacity, high in-plane electrical conductivity, and relatively small volume change on charge/discharge. However, the development of metal chalcogenides for PIBs is still in its infancy because of the limited choice of high-performance electrode materials. However, numerous efforts have been made to conquer this challenge. In this article, we overview potassium storage mechanisms, the technical hurdles, and the optimization strategies for metal chalcogenides and highlight how the adjustment of the crystalline structure and choice of the electrolyte affect the electrochemical performance of metal-chalcogenide-based electrode materials. Other potential potassium-based energy storage systems to which metal chalcogenides can be applied are also discussed. Finally, future research directions focusing on metal chalcogenides for potassium storage are proposed.


| INTRODUCTION
Scientific and technological developments have the potential to affect every part of society. This is particularly the case for lithium-ion batteries (LIBs), which play a role in the daily lives of many people. [1][2][3][4] LIBs have made wearable and portable electronics, electric vehicles, and stationary energy storage possible. Further, they allow the prompt exchange of information without the limitations (such as blackouts) inherent to distributed electricity supply systems (ie, national power grids). [5][6][7][8] From the perspective of environmental protection, LIBs enable the exploitation of intermittent clean, renewable energy sources such as solar, wind, tide, and terrestrial heat, thus allowing the use of fossil fuels to be reduced, reducing environmental contamination, and restricting pollution and waste production to controlled regions or levels. [9][10][11] However, the uneven global distribution of lithium and reserve shortages (20 ppm) have resulted in large increases in market price in the past decade, which Jingwen Zhou and Ye Liu contributed equally to this work. has prevented LIBs from being used in large-scale industries. Thus, sodium and potassium, alkali metal counterparts of lithium, which are present in much higher abundance (23 000 ppm for Na, 17 000 ppm for K) than Li, are popular targets for the development of new rechargeable battery technologies, including sodium-ion batteries (SIBs) and potassium-ion batteries (PIBs), that share a similar "rocking-chair" mechanism with LIBs. [12][13][14] In contrast to the standard redox potential of Na + /Na (−2.71 V vs the standard hydrogen electrode [SHE]), that of K + /K (−2.93 V vs SHE) is much lower and even comparable to the redox potential of Li + /Li (−3.04 V vs SHE), indicating that KIBs could achieve a higher working voltage than SIBs and LIBs. However, K has a larger atomic radius (1.38 Å) than Li (0.68 Å) and Na (0.97 Å) atoms but has the smallest solvated ion radius of these three monovalent alkali metals (3.6 Å for K + , 4.8 Å for Li + , and 4.6 Å for Na + ) in propylene carbonate (PC) solvent because of the weaker Lewis acidity of K + . 13,[15][16][17][18][19] This is a useful feature of K + -based electrolytes, resulting in high ionic conductivity and fast diffusion kinetics. In addition, some computational simulations have even revealed that the diffusion coefficient of K + is nearly four times as much as that of Li + . [20][21][22] As a consequence, graphite, which has been successfully used as the anode material in LIBs, can also be used in PIBs. The intercalation of K + into graphite is favorable and highly reversible, which is a strong contrast to the inactive electrochemistry of SIBs. [23][24][25] Based on the above advantages, the replacement of Li + with K + for the construction of high-performance K-based energy storage systems with enhanced rate capability, high-mass-loading electrodes, and, thus, higher energy densities is highly promising. 13,26 The schematic FIGURE 1 A, Schematic depicting the configuration and working principle of a "rocking-chair" potassium-ion battery. B, A general illustration of the microstructure, merits, and potential applications in the field of potassium storage for metal chalcogenides illustration presented in Figure 1A depicts the general configuration and working principle of "rocking-chair" PIBs.
Of the examples listed above, metal chalcogenides have received numerous insights from researchers in recent years on account of their large theoretical specific capacity, relatively small volume changes, and higher electrical conductivity compared to oxides and phosphides. 53,80 Most metal chalcogenides possess a two-dimensional (2D) layered structure and have low reaction energy barriers for the storage of alkali metal ions. Because of the easy topological extension, numerous techniques can be utilized to prepare different morphologies of metal chalcogenides having the same crystal structure, thus increasing the possible applications of this material. 14,81 In fact, metal chalcogenides are not limited to K + -ion battery applications; these materials are "shining stars," having possible applications in K dual-ion batteries, K-O 2 batteries, K + -ion capacitors, and other energy storage and conversion fields, as shown in Figure 1B. [82][83][84][85][86] In PIBs, depending on the different K + -ion alloying ability of the metal atoms (inactive or active), metal chalcogenides can be divided into two categories: conversion type and conversion/alloying-coupling type. The latter type usually reveals a larger specific capacity because of the contribution to the capacity of the alloying reactions between metal atoms and K + ions.
Even though faster K + -ion reaction kinetics are incorporated into metal-chalcogenide-based PIB anodes, there still remain challenges, unfortunately, including the dissolution and shuttle effect of discharge products (polychalcogenides such as K x S y and K x Se y ), poor ion diffusion in the bulk, and structural instability originating from conversion reactions and agglomeration, which can even result in inactivation caused by volume expansion. Therefore, current metal chalcogenide anodes are still immature and do not deliver satisfactory long-term cycling stability in terms of potassium storage, especially at high rates. 55,80,81 However, an encouraging feature is that, because of the similarity of the energy storage mechanism in metal chalcogenides to that in alkali metal ion batteries, existing modification strategies accumulated for LIBs and SIBs can be potentially transplanted to improve the performance of PIBs. 4,87,88 There are four main approaches to enhance the ionic and electrical conductivity, buffer or accommodate the volume changes, increase the structural integrity and stability, and boost the reaction kinetics of metal chalcogenides ( Figure 2): limitation of discharge depth, nanostructure engineering, confinement by carbon nanophases, or ternary alloying. The majority of these modification methods described above have already been verified to be effective for potassiumbased energy storage according to the results reported previously. 53,81,[89][90][91][92][93][94][95] Furthermore, the interface between the anode and electrolyte in PIBs has drawn recent attention and is another non-negligible factor for stabilizing these energy-storage systems. In particular, the category and component of electrolyte are of great significance to determine whether a thin, homogeneous, and stable solid electrolyte interphase (SEI) with high ionic conductivity is formed on the surface of metal chalcogenide-based anode materials, which results in different structural and surface properties. 62,96,97 The SEI is one of the most crucial features that affect the stable long-term cycling performance of batteries. Simultaneously, optimization to achieve the best match of anode material and electrolyte is also crucial to minimize parasitic reactions and achieve a wide voltage window and large capacity, thus further improving the energy density and working life of PIBs. On the other hand, improving the potassium storage capability of metal-chalcogenide-based materials will not only aid the development of PIBs as high-performance power storage devices but will also facilitate the development of other novel rechargeable systems on the basis of K + -ion storage mechanism (such as K-S/Se batteries, K dual-ion batteries, K + -ion battery capacitors, and K-O 2 batteries) for advanced and practical energy storage applications. [98][99][100][101] In this research progress report, key technical advances and dominated hurdles facing current metal chalcogenide-based PIBs are discussed, with emphasis on how the structure, components, and properties of anode materials synthesized using different strategies affect the electrochemical performance of the assembled PIBs. Meanwhile, the main modification strategies are summarized by exemplifying several previous metal-chalcogenide-based nanostructures and nanocomposites which are particularly designed or prepared. The underlying reasons for the effectiveness of these modifications are proposed as well based on our considerations and reported experimental results. In addition, other representative K + -based novel energy storage systems using metal-chalcogenide-based materials are given to illustrate their great potential for potassium storage further. Finally, we share our insights into the present challenges and possible future research directions, hoping to provide some inspirations for the design and fabrication of highperformance metal chalcogenide-based anode materials for PIBs and other advanced K + -ion energy storage systems.

| METAL CHALCOGENIDE ANODES FOR PIBs
For metal chalcogenides utilized for PIB anodes, they can be mainly classified into two categories according to their different potassium insertion (potassiation) processes:   2.1 | Potassium storage mechanism of metal chalcogenides As mentioned above, metal chalcogenides for PIBs can be divided into two categories according to the reactivity of the metal components with potassium ions.
(1) Conversion-type metal chalcogenides. Representatives of this kind of metal (inactive) chalcogenide are MoS 2 , MoSe 2 , FeS 2 , ZnS, CoSe 1.5 , and V 5 S 8 . 20,90,96,103 The chalcogen component in these compounds is responsible for the vast majority of the capacity, and the inactive metals provide little contribution, except for the part from interfacial storage between metal atoms and chalcogens/potassium polychalcogenides after their reduction to a metallic phase during the potassiation process. Despite their little contribution to the capacity, these metal atoms form a continuously conductive network pinned in the compound, which not only enhances the overall electric conductivity but also serves as a soft substrate to buffer volume changes. The general potassiation/depotassiation reaction mechanism is showninEquation (1).
Here, x and y indicate the stoichiometry of the chemical components of metal chalcogenide, and z is the number of combined/extracted potassium ions reacting with the chalcogens (2 ≤ z ≤ 8, z depends on reaction kinetics, which differ remarkably between materials). M and C represent the metal and chalcogen, respectively. During the initial discharge, the metal (inactive) chalcogenides combine with K + -ions and electrons, becoming transformed into single-phase metal atoms embedded in potassium polychalcogenidess (compounds, xM…K z C). In the first charge process, the xM…K z C compounds can be oxidized by releasing K + ions and electrons to regenerate the metal chalcogen. Of note, on the basis of current experimental results, it is widely recognized that the conversion reactions of metal chalcogenides seem to be not completely reversible in most PIBs. In other words, the phase transformation is partially irreversible, and some new metal chalcogenide phases (with a stoichiometric ratio other than x/y), chalcogens (eg, S and Se) and potassium polychalcogenides (without full depotassiation) are usually formed in association with the conversion of xM…K z CtoM x C y .
Assuming that K z C is the final discharge product, the theoretical specific capacity of conversion-type metal chalcogenides can be calculated using Equation (2).
Here, F is the Faraday constant (96485), and r is the relative molecular mass of the metal (inactive) chalcogenide. The corresponding unit of measurement is milliampere hours per gram (mAh g −1 ).
(2) Conversion/alloying-coupling-type metal chalcogenides. The (active) metal atoms that constitute this kind of metal chalcogenides are commonly metallic elements of groups 14 and 15 with large atomic numbers, such as Ge, Sn, Sb, and Bi, and corresponding metal chalcogenides are represented by Sb 2 S 3 ,S n S 2 ,a n dG e S e . 55,93,104,105 Both metal and chalcogen components are capable of reversible alloying with K + , so they can deliver significantly higher capacities but often suffer from more severe volume expansion during the potassiation process. Hence, this category of metal (active) chalcogenides faces several challenges regarding structural instability. The potassiation/depotassiation mechanisms can be specified in Equation (3).
Here, w is the number of combined/extracted potassium ions reacting with the metal. This is also decided by the reaction kinetics, which is itself affected by the morphology, microstructure, and components of the metal (active) chalcogenides. Similarly, during the initial discharge process, metal-potassium alloy clusters become entangled and potassium polychalcogenides (M x K w …K z C) are produced in the form of aggregates after the full potassiation via conversion and alloying reactions. Upon recharging, K + ions are extracted from the discharge products, resulting in the reappearance of metal (active) chalcogenides. Currently, it is also considered that only a portion of the metal-potassium alloy/polychalcogenides (M x K w …K z C) clusters can return to the original state (M x C y ) based on current cognition, while metal chalcogenides in other new phases, chalcogens, and potassium polychalcogenides are highly possible to exist during the charging process, which is similar to the situation with their inactive counterparts.
Assuming that M x K w and K z C are the final discharge products, the theoretical specific capacity of conversion/ alloying-coupling-type metal chalcogenides can be calculated using Equation (4).
Here, R is the relative molecular mass of the metal (active) chalcogenides.
The detailed K storage mechanisms of these two kinds of metal chalcogenides are exemplified by several impressive works employing ex situ and in situ X-ray (or synchrotron radiation) diffractometry (XRD), Raman spectroscopy, scanning electron microscopy (SEM), and transmission electron microscopy (TEM) as the main phase characterization techniques. For example, Lu's group investigated the potassiation/depotassiation mechanisms of MoSe 2 for the first time by analyzing the ex situ XRD patterns and Raman spectra of the PIB anodes at different states of charge and discharge. 96 After being fully discharged, the MoSe 2 anode materials yielded a diffraction pattern in which the characteristic reflections of MoSe 2 were hard to be detected. In contrast, reflections corresponding to two potassium polychalcogenide species (K 5 Se 3 and K 2 Se 2 ) were observed. More interestingly, in the fully recharged state, elemental Se and Mo 15 Se 19 were easily identified, unlike MoSe 2 , indicating an irreversible part of the recovery from Mo or MoO x to MoSe 2 . Because Se still remains in the system, it inevitably participates in K-Se redox reactions obeying analogous K-Se mechanism that are similar to that in Li-S batteries. The Raman spectra also give a proof of the partial reversibility. Furthermore, the detailed phase transformation of MoSe 2 -Mo/ K x Se-MoSe 2 /Mo 15 Se 19 /Se in one cycle was determined by analyzing XRD patterns obtained at different discharge/ charge voltages. It was deduced that the electrochemical potassiation of MoSe 2 proceeds in two steps: K + intercalation (MoSe 2 +K + + xe − ! K x MoSe 2 , E > 0.53 V vs K + / K) and conversion (K x MoSe 2 + (10 − 3x)K + +(10 − 3x) e − ! 2K 5 Se 3 + 3Mo, E < 0.53 V vs K + /K), and the production of MoSe 2 ,M o 15 Se 19 , and Se mainly occurs in the voltage range of 2 to 3 V upon recharging.
Guan's group investigated the potassium storage mechanism of FeS 2 . 92 The cyclic voltammetry (CV) and XRD results presented in Figure 3A-C illustrate that when the discharge voltage declines to 1.1 V, the peak intensity of the FeS 2 phase is significantly reduced, while a small peak corresponding to KFeS 2 starts to emerge, suggesting the intercalation of K + into FeS 2 interlayers to form K x FeS 2 . After discharging the anode to 0.5 V, only a small peak of Fe phase can be recognized in the XRD pattern. When the electrode was totally discharged, no obvious XRD peaks can be found probably due to the low crystallinity or the tiny particle size in discharge products. However, after the electrode was recharged to 2.8 V, peaks corresponding to KFeS 2 and K 2 Sp h a s e sw e r e mainly observed, indicating that the final charge product is K x FeS 2 . Interestingly, some of the potassium chalcogenides reveal quite a sluggish reaction kinetics, even losing chemical activity on de-alloying. Moreover, the TEM results further prove that Fe and KFeS 2 are embedded in the discharge and charge products, respectively, which is consistent with the XRD analysis. As a result, the group concluded that the initial phase transformation between FeS 2 and K x FeS 2 is an irreversible potassiation reaction, despite the reversibility of the deeper intercalation and conversion stages, as exhibited in Figure 3D.
Guo's group pioneered the exploration of Sb 2 S 3 as an anode material for PIBs. 55 They employed in operando synchrotron XRD (λ = 0.6888 Å) to understand the phase evolution of Sb 2 S 3 during discharge/ charge, as shown in Figure 3E. The shift to smaller diffraction angles associated with Sb 2 S 3 above 0.7 V is evidence for the intercalation of K + , which expands the interlayer distance of Sb 2 S 3 ( Figure 3F). When Sb 2 S 3 was discharged from 0.7 to 0.5 V, diffraction peaks assigned to Sb emerged, indicating the presence of conversion reactions. When the voltage reached 0.1 V, peaks ascribed to K 2 S 6 ,K 2 S 3 ,andK 3 Sb were generated, but they were weak and broad because of the nanocrystallinity. Encouragingly, ex situ selected area electron diffraction (SAED) investigation confirmed t h ep r e s e n c eo fK 2 S 6 and Sb as intermediate products on the basis of their polycrystalline diffraction rings. Thus, the potassiation mechanism for Sb 2 S 3 can be summarized by Equations (5)-(7).
Whereas the phase evolution during charging was not involved in the literature, aside from the confirmation of the presence of undecomposed K 3 Sb and K 2 S 3 in the recharged products. In addition, Guo et al proposed that the continuous pulverization of Sb 2 S 3 particles caused by the huge volume change (ca. 300%) during cycling is responsible for the rapid decay in capacity. In addition, Xia's group investigated the potassiation/depotassiation behavior of SnS 2 using ex situ XRD, TEM, and SAED measurements. 93 According to the XRD results, the conversion reaction of SnS 2 begins above 1.0 V, as shown by the presence of diffraction peaks corresponding to SnS, K 2 S 5 ,a n d K 2 S. Subsequently, further conversion and alloying reactions occur, and the dominant capacity contribution occurs between 0.01 and 0.5 V. The final discharge products were determined to be KSn and K 2 S (primary) and α-Sn and K 4 Sn 23 (minor). However, even though the electrode was recharged to 2.0 V, only SnS coupled with K 2 S( i n s t e a do f SnS 2 ) could be found. These findings were confirmed by the TEM analysis of the fully discharged/charged electrodes.
These aforementioned works sufficiently reveal that, for metal-chalcogenide-based PIB anodes, the potassiation FIGURE 3 A, Cyclic voltammetry (CV) curves; B, initial discharge/charge curves and C, ex situ X-ray diffractometry (XRD) patterns at various states of pure FeS 2 anode. D, Schematic illustration for its potassium storage reaction mechanism. Reproduced with permission from Reference 92 Copyright 2018, Wiley-VCH. E, In situ synchrotron XRD patterns of Sb 2 S 3 upon K insertion (left) and ex situ selected area electron diffraction (SAED) pattern (right) at 0.5 V. F, Image plots of the (212) reflection and corresponding fitted peak. Reproduced with permission from Reference 55 Copyright 2018, Nature Publication Group and depotassiation depth intimately relies on their morphology, microstructure, components, and interfacial bonding, and phase transformation is very likely to be incompletely reversible. In addition, the decomposition kinetics for potassium polychalcogenides is commonly not fast enough, resulting in remnant K z Ci nt h ef u l l y recharged electrodes.

| Challenges facing metal chalcogenides
Even though, a few metal chalcogenides have been proven to be electrochemically active when employed for potassium storage, it does not mean that there are no scientific issues and technical hurdles before discussing the feasibility of metal chalcogenides for practical PIBs and other K-based energy storage systems. 13,15,116 These challenges facing current metal chalcogenide anodes can be mainly concluded as the following aspects: 1. In most metal chalcogenides (mainly sulfides/selenides), potassium polychalcogenides are generated during the deep potassiation process (discharge voltage cutoff: 0.005 V), which may result in detrimental shuttle effects similar to those observed in Li-S batteries. The parasitic reactions between highly reductive potassium metal (in PIB half-cells or other systems with potassium metal anodes) and potassium polychalcogenides produce a K 2 C film on the metal anode surface, which blocks the channels for K + -ion diffusion, thus dramatically decreasing the capacity within the initial cycles and even shutting down the battery suddenly. This is also a potential disadvantage for PIB full cells, which do not employ potassium metal as the anode, because cathode materials commonly exhibit a certain degree of reducibility. Simultaneously, affected by the above unfavorable factors, the evaluation of the potassium storage capability via half-cell methods imposes further burdens on matching the capacity of anodes and cathodes for fullcell PIBs. 2. The huge volume change and incompletely reversible phase transformation during discharge/charge cycles lead to serious structural instability in metal chalcogenides. Obviously, On the one hand, incompletely reversible redox reactions result in the loss of some active components during the depotassiation processes, and, hence, the reversible specific capacity is gradually downward in the following cycles. On the other hand, the tremendous volume change is the chief culprit of the pulverization of active materials and even the exfoliation from the current collectors.
This results in the accumulation of electrochemically inert components in the electrode. This is the dominant reason why most present metal chalcogenides demonstrate unsatisfactory cycling stability. 3. The intrinsically limited electric and ionic conductivities hamper both electron transport and K + -ion diffusion in bulk metal chalcogenides, leading to a sluggish reaction kinetics, as well as the low utilization efficiency of active sites. As a result, large electrochemical polarization and remarkable capacity decay are easily observed in the measurements at high rates. 4. The instability of SEI is another difficulty that needs to be addressed urgently for metal chalcogenides. The significant volume expansion and shrinkage during cycling destroys the generated SEI film in the initial stages of use, and the newly exposed metal chalcogenides react with K + ions to form a new SEI coating. The repeated generation and destruction of the SEI results in the excess consumption of K + ions, thereby lowering the CE. Simultaneously, due to the high activity of potassium salts, traditional electrolyte components for LIBs and SIBs seem to be not so much effective for PIBs. Although a dozen combinations of potassium salts and solvents have been attempted as electrolytes for PIBs, it still remains challenging to construct an ideal SEI film having high ionic conductivity, sufficient mechanical ductility and strength, and satisfactory chemical stability, especially for certain categories of anode materials.

| Optimization strategies for metal chalcogenides
On the way to pursuing suitable metal chalcogenides for PIBs with excellent long-term cycling stability at high rates, numerous efforts have been devoted to materials modification. Benefitting from the long history of research into LIBs, quite a few metal-chalcogenide-based nanostructures and nanocomposites having enhanced potassium storage capabilities have been designed and synthesized. For convenient comparison, the electrochemical performance of typically modified metal chalcogenides for PIBs has been concluded in Figure 4A. 20,53,55,56,80,81,[89][90][91][92][93][94][95][96][97]111,112,115 As shown, there is still large room for improvement, especially concerning the large capacity, low discharge voltage, and long-term cycling stability. Currently, there are four main modification strategies: limitation of discharge depth, nanostructure engineering, confinement by carbon nanophases, and ternary alloying. In the following section, we discuss the advantageous and disadvantageous features of each strategy are exemplified using recent impressive works.
Besides, we also propose a road map for the future development of metal chalcogenides before the discussing. As shown in Figure 4B, we define two conceptual promotion strategies: "inner design" and "outer design." Inner design involves the selection of the most favorable electrolyte components and cycling conditions for certain kinds of metal chalcogenides, thus obtaining the most adaptable couples, whereas outer design involves changing the material's properties or constructing/manipulating the SEI, thus overcoming the intrinsic weakness of metal chalcogenides by artificial solutions. These two strategies will have significantly positive effects on the electrochemical performance of PIB anodes. 117,118 Meanwhile, computational simulations 119,120 and in situ characterization 121,122 will help us understand the underlying potassiation/depotassiation mechanisms and, thus, develop metal chalcogenides for high-performance PIB anodes that are suitable for industrial production. 123 Of course, better performance implies the presence of more compatible interfaces (anode/SEI and SEI/electrolyte, of which the structures, components, and mechanisms are unclear at the present stage of research) in PIBs, and investigation in this area will in turn, accelerate the advance of system design. 124 For most conversion-type metal (inactive) chalcogenides, there is usually an intercalation reaction before the conversion reaction during discharge. The corresponding intercalation mechanism is similar to the mode of intercalating K + ions into graphite. 108,126 Therefore, relatively small volume expansion occurs at this stage of potassiation and the structural integrity is well maintained. Because there are no conversion reactions, the K + -ion insertion and extraction kinetics are much faster and the reversibility of the phase transformation should be greater.
For example, Wu's group employed a 2D layered MoS 2 as a host material for potassium ion intercalation within the electrochemical window of 0.5 to 2 V. 126 According to CV results and XRD characterizations, a stable stoichiometric K + -ion intercalation compound, K 0.4 MoS 2 , was found to store potassium through highly reversible K + -ion intercalation and de-intercalation reactions in an electrolyte comprising 0.5 M KPF 6 dissolved in PC/ethylene carbonate (EC) (volume ratio [V PC / V EC ] = 1:1). This compound generated in the first stage of potassium intercalation (around 1.02 V), and further intercalation at ca. 0.8 V leads to the formation of K x MoS 2 (x > 0.4), but the continuous accumulation of this material induces the degradation of MoS 2 to potassium sulfides (K x S), resulting in the loss of structural integrity of MoS 2 . Thus, the discharge depth was set to 0.5 V to ensure that K 0.4 MoS 2 is the only discharge product for participation in the subsequent redox reactions.
Unlike MoS 2 ,K 0.4 MoS 2 exhibits a quite limited expansion of 34% along the c-axis in the potassiated state, even less than K + -ion-intercalated graphite (ca. 67%). The excellent preservation of the structure endows this compound with a very steady cycling behavior; namely, its reversible specific capacity of 63.8 mAh g −1 is retained after 200 cycles with a capacity retention of 97.5%. Further, the discharge/charge profiles almost overlapped during cycling, and the corresponding CEs remained above 99.2%, well illustrating the high reversibility. In addition, Shu's group reported that pretreated commercial TiSe 2 could be used as an insertion-type anode for PIBs. 89 They employed in situ XRD measurements to investigate the potassiation and depotassiation mechanisms of TiSe 2 in the voltage range of 1.0 to 3.0 V ( Figure 5A,B). The (002) reflections were unrecoverable, and this was ascribed to the insertion of K + ions into the vacant spaces between the van der Waals layers of TiSe 2 , resulting in the separation of TiSe 2 and phase transformation instead of lattice strain in the first cycle. The first discharge process to 1.0 V resulted in the formation of K 0.8 TiSe 2 , and, subsequently, K 0.24 TiSe 2 was produced after full recharging to 3.0 V, resulting in some capacity loss during the incomplete depotassiation reaction. This newly formed K 0.24 TiSe 2 phase has been shown to serve as a stable K + -ionintercalating host material for reversible potassium storage in an electrolyte comprising 0.8 M KPF 6 in diethylcarbonate (DEC) and EC (V DEC /V EC = 1:1). The oscillation between K 0.24 TiSe 2 and K 0.8 TiSe 2 proceeds in the subsequent cycles, as shown in Figure 5C,D. Fortunately, fast reaction kinetics (72.6 mAh g −1 at 0.75C and 38.3 mAh g −1 at 16C) coupled with decent cycling stability (300 cycles at 2C) were synchronously achieved in this case ( Figure 5E,F). After solving the irreversible capacity loss by using pretreatment cycles, TiSe 2 (K 0.24 TiSe 2 , more precisely) was matched with pretreated Prussian blue to construct a full-cell PIB with high initial CE (ICE) and stable cycling performance.
However, it is noteworthy that the cycling stability is improved at the sacrifice of the specific capacity under this circumstance. Approximately at least two-third of the capacity for potassium storage arises from the conversion and alloying reactions of the metal chalcogenides. 103,108 Thus, limiting the discharge depth to intercalation reaction cannot take advantage of this dominant portion of the theoretical capacity. It also needs to point out that the relatively high discharge voltage plateau of intercalation reactions will reduce the working voltage and energy density of assembled full-cell PIBs.

| Nanostructure engineering
Considering that the intrinsically low electric and ionic conductivities in the bulk are already one of the most critical hindrances to the use of metal chalcogenides as efficient potassium ion host materials at high rates, nanostructure engineering should be an effective strategy to promote their reaction kinetics by quantum confinement effects, which means a distinctly higher utilization of active materials and a larger potassiation/depotassiation depth. 2,7,127 Besides, constructing nanostructures not only helps shorten ion diffusion pathways and facilitate electron transfer, but also self-accommodates volume change during discharge and recharge cycles, alleviating the electrochemical collapse caused by mechanical stress. As a result, the potassium storage capability of nanostructured metal chalcogenides has been significantly enhanced, especially in terms of capacity and cycling performance.
Wang's group has published several interesting works related to this research direction. First, they prepared a flexible freestanding anode consisting of metallic octahedral CoSe 2 threaded with N-doped carbon nanotubes for PIBs through solvothermal growth and selenylation ( Figure 6A). 20 These octahedral CoSe 2 nanoparticles with an average size of 150 to 200 nm are highly dispersed along the carbon tubes in sequence, thus providing sufficient void space to buffer the volume expansion upon potassiation and improve the structural stability during cycling ( Figure 6B,C). Benefiting from the synergistic effect of the fast electron transfer of nanosized CoSe 2 itself and the unique hierarchical structural advantages of the composite, this freestanding CoSe 2 -based framework demonstrated superior electrochemical performance when tested as a PIB anode in 0.8 M KPF 6 +DEC/EC (vol/vol = 1:1) electrolyte. A stable reversible capacity as high as 296 mAh g −1 w a so b t a i n e dw i t ht h i sh y b r i da n o d e at 0.2 A g −1 , and there was more than a two-third capacity contribution at voltages below 0.5 V ( Figure 6D). Further, when the current density was increased from 0.2 to 2 A g −1 , the material exhibited a high capacity retention of ca. 68%. What is more attractive is its excellent long-term cycling stability. The electrode could be operated steadily for 600 cycles at 2 A g −1 and showed negligible capacity loss even after the last cycle ( Figure 6E). Furthermore, the phase transformation of CoSe 2 -Co 3 Se 4 -(Co 3 KSe 4 +CoSe)-(Co+K 2 Se) explained as the detailed potassiation procedure via firstprinciple calculations. Next, a unique nanostructure of bamboo-like hollow MoS 2 /N-doped C tubes, as displayed in Figure 6F, was reported by their group. 90 These hybrid tubes mainly consist of MoS 2 (ca. 75 wt%). Interestingly, the distance between the adjacent (002) planes of MoS 2 reached 10 Å in the tubes because of the presence of carbon interlayers, proven by Figure 6G,H. The expanded MoS 2 layers with abundant defects could provide more active sites for potassium storage, as well as more channels for ion diffusion, which is significant for reducing the ion migration energy barrier and then improving reaction kinetics. The special 1D hollow microstructure endows the composite with high stability, given that the mechanical stress can be gently released because the volume change is along the direction of tube walls. The high capacity of 451 mAh g −1 coupled with outstanding cycling stability (74% capacity retention after 1000 cycles) is evidence of these favorable features ( Figure 6I,J). It is worth noting that, even when the scan rate was increased to 1.1 mV s −1 , faradaic pseudocapacitance contributed to 73% of the capacity, which is unlike the situation in carbon-based materials (>90%). Density functional theory calculations were used to illustrate the merits of K + -ion adsorption in MoS 2 with N-doped C interlayers compared to bulk MoS 2 .Af u r t h e rm o d i f i c a t i o nt o construct 3D interconnected MoS 2 networks has been carried out by Wang's group. 53 T h en a n o s i z e dr o s e -l i k eM o S 2 anchored on reduced graphene oxide (rGO) was easily fabricated by a one-step hydrothermal method. The chemical Mo-C and Mo-O-C bonds guarantee the intimate interfacial connection between MoS 2 and rGO. Simultaneously, the 3D interconnected networks provide abundant electron transfer and ion migration channels and alleviate the agglomeration of the 2D nanosheets caused by van der Waals interactions during cycling. The above features give this nanosized rose-like MoS 2 outstanding rate and cycling behavior. More surprisingly, only 46.2% of the capacity is contributed by pseudocapacitance even at 1 mV s −1 ; thus, the high capacity retention at high rates is mainly achieved by the diffusion-controlled bulk conversion r e a c t i o n s ,w h i c hp r o v i d eam u c hl a r g e rc o n t r i b u t i o nt o the capacity than surface pseudocapacitance.
Of course, it also deserves to highlight several other representative works in this research area. Guo's group utilized hollow carbon spheres as a hard template to anchor few-layered V 5 S 8 nanosheets directly on the surface through solvothermal growth and a subsequent phase transformation step via high-temperature annealing. 81 On the one hand, the 2D ultrathin structure greatly shortens the ion transport pathway, enabling a faster conversion reaction. On the other hand, ultrathin V 5 S 8 nanosheets can be tightly pinned on the carbon nanospheres, thus self-confining the volume change, enhancing the structural stability, and preventing the aggregation of nanosheets. Hence, a high reversible capacity of 645 mAh g −1 was achieved at 0.05 A g −1 , and around 190 mAh g −1 was retained after 1000 cycles at 2Ag −1 without obvious capacity decay compared to the initial cycle. In addition, the capacity of 153 mAh g −1 at 10 A g −1 achieved by V 5 S 8 /carbon hollow spheres is the best rate capability among all reported metal-chalcogenide-based PIB anodes to date.
The original investigation of Sb 2 S 3 nanosheets for PIBs was also conducted by Guo's group. 55 They produced few-layered Sb 2 S 3 nanosheets with an amorphous carbon coating via solution-triggered one-step shear exfoliation of bulk Sb 2 S 3 in water/ethanol solvent. The asprepared Sb 2 S 3 nanosheets delivered a large reversible capacity of approximately 500 mAh g −1 at 0.05 A g −1 and demonstrated better cycling performance in KSiF 6 + EC/ PC (vol/vol = 1:1) electrolyte than bulk Sb 2 S 3 . The main reason for the difference lies in the fact that the soft 2D microstructure of the Sb 2 S 3 nanosheets not only boosts ion diffusion along short pathways but also enables the reversible recovery to the original structure by buffering and self-accommodation of the volume changes, whereas bulk Sb 2 S 3 is easy to suffer from cracking and pulverization during cycling. In addition, Lu's group synthesized a flower-like nanostructure assembled from MoSe 2 /Ndoped carbon nanosheets (MoSe 2 /N-C) 96 in which only a small number of amorphous N-doped carbon layers are attached to MoSe 2 nanosheets. In contrast to pure MoSe 2 blocks, MoSe 2 /N-C shows an obviously higher and more stable capacity during cycling, demonstrating the promise of nanostructure engineering. Xia's group reported a novel kind of PIB anode material comprising few-layered SnS 2 supported on rGO. 93 Similarly, the nanostructured SnS 2 architecture delivered a higher reversible capacity of 448 mAh g −1 and longer working lifetime of up to 300 cycles. It should be emphasized that nearly 73% (325 mAh g −1 ) of the entire capacity is obtained in the voltage range of 0.01 to 0.5 V, which is more superior to those of most previously reported anode materials. Thus, this result demonstrates a significant step toward highenergy-density full-cell PIBs to some extent.
Although the aforementioned works reveal the positive aspects of nanostructure engineering, some unfavorable influence is worth our reconsideration as well. First, it is recognized that nanosized active materials have a low tap density. This is a critical performance index for industrial-grade products, and a low tap density makes it hard to guarantee the volumetric energy density of the electrode. Second, nanostructured materials expose more surface areas while having higher chemical activity at the same time. Therefore, how to generate a thin, stable, mechanically strong, and continuous SEI film with high ionic conductivity on the material surface remains a significant challenge. Thirdly, a larger SEI will mean greater consumption of K + , thus increasing the irreversible capacity and reducing the ICE if without other modification techniques. Finally, aside from the buffering effect, the problem of structural instability has not been addressed thoroughly. In particular, metal-chalcogenidebased nanostructures always lack a physical or chemical barrier to prevent the pulverization of active components.

| Confinement by carbon nanophases
Because of its high structural stability and lightweight, carbon nanophases have already been combined with metal chalcogenides to prepare novel composites for applications requiring high structural stability. Fortunately, this traditional modification strategy, which has been used for LIBs and SIBs, is also effective for PIBs and allows the improvement of the electrochemical performance. On the one hand, carbon nanophases can play a role of soft barrier to confine the volume changes and self-aggregation of the metal chalcogenides during potassiation and depotassiation, significantly enhancing the tolerance to structural damage. On the other hand, the carbon matrix is also able to supply high-speed electron transfer pathways to the dispersed metal chalcogenides, thus improving the electrical conductivity of the composites. Two forms of confinement by carbon nanophases have been used for metal chalcogenides. In one, tiny metal chalcogenides are embedded into carbon matrix directly, and the volume expansion is restricted by the internal stress of carbon phase. The other is to encapsulate metal chalcogenides in carbon nanophases with void spaces, allowing controllable volume expansion within a restricted region.
Obviously, the former method is easier to realize in terms of material preparation. Electrospinning is widely used to construct composites of metal chalcogenides embedded in carbon. For example, Zhang's group prepared graphene-encapsulated FeS 2 nanoparticles embedded in carbon nanofibers (FeS 2 @G@CNF) through electrospinning and calcination. 80 The introduction of graphene facilitated the uniform distribution of ultrafine FeS 2 nanoparticles in the carbon nanofibers, and there was little FeS 2 at the surface zone of the nanofibers. When evaluated as the anode material for PIBs, FeS 2 @G@CNF exhibited acceptable cycling stability over 680 cycles (ca. 62% capacity retention), even though some capacity decay was observed within the initial 100 cycles. Qian's group used a similar strategy to prepare ultrafine VSe 1.5 nanoparticles embedded in carbon nanofibers. 95 In this composite, the VSe 1.5 nanoparticles are closely surrounded by disordered carbon, which effectively confines the migration or aggregation of VSe 1.5 during cycling, as shown in Figure 7A,B. Although only the data from the initial 100 cycles was provided, the capacity was maintained at around 177 mAh g −1 at 1 A g −1 ( Figure 7C). Zhang's group specially designed 3D amorphous carboncoated CoS/N-doped CNTs anchored on CoS-coated carbon nanofibers as an anode material for PIBs. The composite was prepared by electrospinning, seed-assisted growth, in situ CVD, and high-temperature sulfuration. 94 Some of the CoS nanoparticles with an average diameter of ca. 20 nm were encapsulated at the ends of the CNTs, whereas the remainder was anchored on the surface of the amorphous N-doped carbon nanofibers. Benefiting from confinement effects, this electrode demonstrates a stable capacity of nearly 400 mAh g −1 and decent cycling stability of up to 600 cycles at 3.2 A g −1 . However, the material showed severe capacity loss in the first 100 cycles, possibly In contrast, the latter confinement strategy using nanocarbon with void spaces seems to be more efficient in maintaining the structural stability of composites during potassiation and depotassiation cycles. However, the microstructure and corresponding preparation procedure requires innovative and careful design. The yolk-shell structure is one of the most promising candidates in this kind of modification strategy, providing suitable void space for metal chalcogenides to accommodate their volume expansion, but this expansion is seriously restricted in the "shell"-protected area. Thus, very little mechanical stress caused by volume expansion will be imposed on the carbon shell, and the pulverization of metal chalcogenides will be significantly alleviated within the nanoconfined space. Thus, the structural stability coupled with ion/electron transfer kinetics of the whole composite are highly improved. An impressive work reported by Guan's group is worth highlighting here. 92 They synthesized a yolk-shell FeS 2 @C nanostructure anchored on a graphene matrix. The controllable volume change of FeS 2 happens insidethecarbonshellwhiletheintercrossedgrapheneconnects those yolk-shell FeS 2 @C units to ensure the high conductivity and sufficient structural stability (Figure 7D,E,G). When tested in 1 M KPF 6 +EC/PC(vol/vol=1:1)electrolyte, the composite delivered a stable capacity of around 100 mAh g −1 over 1500 deep cycles at a high rate of 5 A g −1 ( Figure 7F,H). Guan's group also prepared a 0D/2D hybrid structure comprising yolk-shell NiS x @C embedded in carbon nanosheets (Y-S NiS x @C). 111 In this composite, numerous NiS x @C and carbon nanocage units with an average size of ca. 30 nm were homogeneously embedded in a carbon nanosheet matrix. Under the double protection of the carbon "shell" and carbon nanosheets, the Y-S NiS x @C achieved an ultralong cycling stability of 8000 cycles, and, surprisingly, the capacity was maintained at approximately 130 mAh g −1 .F u r t h e rmore, the unique microstructure was well maintained after cycling tests. Not limited to this, Bao's group designed another double protective structure for ZnS, where hybrid microspheres assembled from ZnS nanorods and an amorphous carbon coating were sandwiched between rGO layers. 56 There are enough voids between the ZnS/C microspheres and rGO substrate to buffer the volume expansion of ZnS. The amorphous carbon coating acts as the first protective layer, holding ZnS nanorods together like binders, whereas the second protective rGO layer limits the aggregation or pulverization of the ZnS/C microspheres and offers fast pathways for electron transfer, as illustrated in Figure 7I,J. Hence, stable cycling behavior for over 300 cycles was obtained, after some initial activation cycles, as displayed in Figure 7K.
Consequently, confinement using carbon nanophases shows better feasibility in improving the electrochemical performance of metal chalcogenides, especially the longterm cycling stability at high rates. It is still worth noting that the excess consumption of K + ions caused by SEI formation leads to an obvious decrease of ICE, on account of the introduction of porous or amorphous carbon nanophases. Moreover, metal chalcogenides commonly exist in carbon nanophases in the form of nanosized structures; thus, an inadequate tap density is another issue of concern.

| Ternary alloying
The ternary alloying method is a new concept for metal chalcogenides modification proposed in recent years. It can take care of both the aforementioned problems concerning tap density and electrochemical performance. The main aim of this modification is to enrich the material with vacancies, adjust the electronic structure, and optimize the physical properties of metal chalcogenides by introducing a third-party cation (metal atom) or anion (chalcogen), which helps to increase the number of active sites, facilitate reaction kinetics, and alleviate the structural vibrations of ternary metal chalcogenides.
Guo's group has pioneered the design, synthesis, and structural/electrochemical investigations of ternary metal chalcogenides that are suitable for PIB anodes. They fabricated MoS 2(1-x) Se 2x alloys (x = 1, 0.75, 0.5, 0.25, and 0) by replacing Se atoms with S atoms in MoSe 2 nanoplates via a facile alloying approach. 112 They found that an S/Se ratio of nearly 1:1 endows the alloys with the largest vacancy concentration. These vacancy sites not only boost electron transfer but reduce the K + -ion adsorption energy barrier and alleviate structural change during cycling ( Figure 8A-D). When used as a PIB anode material, the MoSSe alloy delivered a high reversible capacity of 517.4 mAh g −1 at 0.1 A g −1 and good cycling stability (220.5 mAh g −1 after 1000 cycles at 2 A g −1 ), as shown by the electrochemical performance illustrated in Figure 8E-G. They also prepared another ternary metal chalcogenide (SnSb 2 Te 4 ) through the substitution of cations. 97 To improve the electric and ionic conductivity of the SbSb 2 Te 4 p-type topological insulator, nanodots were dispersedinafew-layeredgraphene substrate by a facile ball-milling method. The reaction kinetics is accelerated by the PN heterojunction and the conductive topological surface, and excellent potassium storage was achieved in 1 M potassium bis(fluorosulfonyl)imide (KFSI) + DMC electrolyte. Furthermore, Wang's group also reported few-layered ternary chalcogenide Ta 2 NiSe 5 nanosheets for PIBs. 91 Ion-intercalation-mediated exfoliation yields Ta 2 NiSe 5 nanosheets with an expanded interlayer distance of 1.1 nm, as well as abundant Se sites originating from the NiSe 4 and TaSe 6 octahedral clusters. The above features are favorable for potassium storage so that they can demonstrate a large capacity of 315 mAh g −1 with a moderate ICE of 70%, decent rate capability (121 mAh g −1 at 1 A g −1 ) and stable cyclability (81.4% capacity retention after 1100 cycles).
Thus, research to date indicates that ternary alloying is a very promising modification strategy for metal chalcogenides on basis of known research findings. However, current investigation into the alloy categories, physical structures, and electrochemical mechanisms is far from adequate. It is believed that more suitable ternary metal chalcogenides for high-performance PIBs will be developed if more insights are revealed in this direction.

| Regulating the electrolyte for metal chalcogenides
Compared with electrode material modification, optimization of the electrolyte has drawn less attention from researchers in PIB research. However, due to the high activity of K + -ion-based electrolytes, the appropriate pairing of potassium salt and solvent is a critical step for constructing moderate SEI layers, reducing parasitic reactions, and maintaining structural stability. However, research into this topic began late and is still in infancy. Despite several kinds of electrolytes have already been used for PIBs (mainly KPF 6 , KFSI, or potassium bis(trifluoromethane) sulfonimide as solutes and employing EC/DEC, EC/PC, or dimethoxyethane [DME] as solvents), but their effects on the electrochemical performance of a certain metal chalcogenide have usually been neglected. To date, very few attempts at optimizing the electrolyte component of  96,97 For example, Lu's group studied the effect of four kinds of electrolytes on the cycling behaviors of MoSe 2 nanosheets. 37 When a 1 M KPF 6 + DME electrolyte was employed, the battery stopped functioning suddenly in the 13th cycle. In the cases of 1 M KPF 6 + EC/DEC (vol: vol = 1:1) and 1 M KPF 6 + EC/DMC (vol:vol = 1:1) electrolytes, rapid capacity decay occurred over the first 100 cycles. In contrast, MoSe 2 nanosheets maintained stable cycling in 1 M KFSI + EMC and 3 M KFSI + DME electrolytes, but the reversible capacity was slightly higher in the former electrolyte. The authors hypothesized that the more stable SEI formed by KSFI is responsible for the long cycling stability ( Figure 9A). Guo's group investigated the electrochemical performance of SnSb 2 Te 4 nanodots in KFSI + DMC or KSFI + EC/DEC electrolytes. 97 The SnSb 2 Te 4 nanodots delivered very stable capacities of around 320 mAh g −1 over 200 cycles in the KFSI + DMC electrolyte, whereas an obvious capacity decrease began after the 120th cycle in the KFSI + EC/DEC electrolyte. The galvanostatic cycling profiles of symmetric foil cells in different electrolytes indicate that K foil shows the lowest voltage hysteresis in the KFSI + DMC electrolyte ( Figure 9B,C). The stability of the K foil is regarded as the main reason for the longer cycling lifetime.
The above attempts illustrate the important role of the electrolyte in achieving high-performance metal-chalcogenide-based PIB anodes. However, the underlying mechanisms and general design principles need more detailed explorations to specify.

| METAL CHALCOGENIDES FOR POTENTIAL POTASSIUM STORAGE SYSTEMS
In addition to the widely investigated PIBs, potassiumsulfur/selenide batteries, potassium-ion capacitors, and potassium-oxygen batteries have emerged as novel types of potassium-based energy storage media due to the relatively high energy densities. [128][129][130] Recently, the intriguing electronic properties of metal chalcogenides make this category of materials available for the above applications. In this section, we briefly discuss the progress and

| Cathode stabilizers for K-S/Se batteries
Unlike the fundamental intercalation reactions in Li/Na/ K + -ion batteries, potassium-sulfur (K-S) and potassiumselenide (K-Se) batteries are based on the conversion or integration of potassium polysulfide/selenide analogs on the cathode. The performance of K-S/Se batteries is severely inhibited by three main factors as follows 128,131 : 1. The dissolution of potassium polysulfide/selenides and the shuttle effect, which result in the accumulation of potassium sulfides/selenides (obtained through the reduction of potassium polysulfide/selenides by highly reductive potassium) on the metal anode surface and blockage of K + -ion diffusion channels from electrode to electrolyte. 2. The low utilization of the theoretical capacity caused by the weak electric and ionic conductivity of elemental sulfur/selenium. 3. Structural instability toward pulverization because of the large volume expansion and shrinkage of sulfur/ selenium during repetitive potassiation and depotassiation processes. 4. The severe side reactions originating from the highly reactive potassium metal anodes.
At the present stage, efforts have been mainly focused on addressing the above issues facing sulfur/selenium cathodes according to recent reports. Carbon substrates with porous structures are mainly utilized as host materials to improve the ionic/electric conductivity of S/Se, as well as buffering their volume expansion upon discharging to enhance the structural stability. For example, sulfur can be stored in microporous carbon substrates via impregnation or vaporization-condensation methods. Xu's group has carried out an impressive work related to this. 100 In their studies, small allotropes of sulfur (S n , n ≤ 3) were confined in a microporous carbon matrix, which prevented the sublimation and decomposition of sulfur via the strong interactions between the fragment sulfur species and carbon atoms, thereby effectively alleviating the solvation of polysulfides and enhancing the reaction kinetics. This synergistic effect endowed this K-S battery with a reversible capacity of 1198.3 mAh g −1 , moderate CE of 97%, and good cycling stability, retaining 869.9 mAh g −1 after 150 cycles. By impregnating SeS 2 into an N-doped freestanding porous carbon nanofiber film, a self-supported SeS 2 /C cathode was fabricated by Yu's group. 129 The encapsulation of nanosized SeS 2 into a carbon matrix not only minimized the volume expansion but shortened the ion transport pathways. Meanwhile, these interconnected carbon nanofibers also provide a highly conductive network for electron transfer. As a result of this synergistic effect, the electrode had an enhanced rate capability and extended cycling lifespan. Conductive polymers have also been used as host substrates for S/Se storage. Recently, Sun's group found that sulfurized polyacrylonitrile (SPAN) composites are promising cathodes for K-S batteries with polyacrylic acid (PAA) as a binder. 131 The PAA greatly improved the structural stability of the SPAN electrode because of the excellent cohesion of the active materials with the electrode, which increased its ability to accommodate the large volume expansion upon cycling. As a result, a high reversible capacity of 1050 mAh g −1 and good reversibility (95% capacity retention after 100 cycles) were achieved.
Whereas the lack of physical barrier around the electrolyte means that polysulfide/selenides can still be extracted from the host materials, dissolved in the electrolyte, and diffuse into the potassium anode, leading to little improvement in the cycling stability, as proved by the above works. Awareness of this issue has increased recently, and the construction of a protective layer around the active materials has been proposed as a solution. Chen's group used a coating of conductive polyaniline (PANI) on a CMK-3/sulfur hybrid cathode, 132 and this was found to promote the cycling performance significantly (62% capacity loss for pure sulfur compared to 28% for PANI-coated CMK-3/sulfur hybrid). Although the structure of the PANI-coated cathode was well preserved after 50 cycles, a few long-chain polysulfides were still shuttled to the metal side. Obviously, developing efficient physical barriers for advanced K-S/Se batteries is one of the most urgent tasks on the way to advanced K-S/Se batteries.
Inspired by the excellent protection against and adsorption of polysulfide/selenides by metal chalcogenides, we propose the possibility of them to be as efficient cathode stabilizers for K-S/Se batteries. This consideration is partially based on related works involving Li-S/Se and Na-S/Se batteries. For example, Jin's group synthesized Co 9 S 8 nanocrystals with inlaid carbon (Co 9 S 8 /C) hollow nanopolyhedra derived from metal-organic frameworks using an impregnation method as a confinement barrier to host sulfur (S@Co 9 S 8 /C). 133 On the one hand, the large inner space of the Co 9 S 8 /C hollow nanopolyhedra ensured sufficient mass loading of S and accommodated the volume expansion of the Li 2 S x species. On the other hand, the polar Co 9 S 8 /C shells provided spatial confinement and chemical bonding to restrict the shuttling of lithium polysulfides. As a result of these favorable factors, the S@Co 9 S 8 /C cathode yielded a high reversible capacity and very small capacity decay of 0.041% per cycle over 1000 cycles at 2C. Also conducted by Jin's group was interlaced carbon nanotubes threaded around hollow Co 3 S 4 nanoboxes (CNT/Co 3 S 4 -NB) prepared by hydrothermal sulfuration. 134 The synergistic effect of ultrahigh charge transfer properties and physical/chemical confinement of Co 3 S 4 nanoboxes meant that CNT/Co 3 S 4 -NB functioned as an efficient cathode stabilizer for Li-S batteries, which exhibited significant improvement in the electrochemical performance, especially in terms of long-term cycling stability. Metal chalcogenides are also capable of cathode stabilizers and catalysts in Na-S batteries. We specify this positive function by exemplifying two representative works employing MoS 2 and NiS 2 , respectively. Xu's group prepared an S/MoS 2 /C hybrid cathode by growing MoS 2 nanosheets on hollow carbon spheres as the S carrier for Na-S batteries. 135 The hollow carbon spheres provide space to host sulfur and improve its conductivity, and the MoS 2 nanosheets can effectively adsorb the soluble sodium polysulfide intermediates and alleviate the shuttle effect. Consequently, the cycling stability of this Na-S battery was extended to 1000 cycles. Dou's group synthesized N-doped CNTs implanted with NiS 2 crystals as a multifunctional sulfur host. 136 The polarized NiS 2 molecules not only immobilized the polysulfides by chemical adsorption but also facilitated the conversion of sodium polysulfides to Na 2 S on discharging and S 8 on charging, as supported by the in situ synchrotron XRD results. These favorable features increased the battery's rate capability and lifespan to 750 cycles.
Although metal chalcogenides have not been utilized to host or stabilize potassium polychalcogenides in K-S/ Se batteries until now to our best knowledge, it is believed that they can be easily extended into this field on account of the fact that their polar metal-chalcogen bonds and dangling chalcogen sites can efficiently adsorb polysulfides and accelerate the phase transformations in reported Li/Na-S batteries, as discussed above, thus greatly enhancing the specific capacity and cycling stability. Theoretically speaking, this stabilization strategy is considered to be equally effective in potassium storage systems such as K-S/Se batteries, which share a similar working mechanism.

| Anode materials for potassium dual-ion batteries
The dual-ion battery (DIB) is a novel kind of energy storage system involving the separate intercalation of anions and cations into the electrodes during the charge/discharge process. The high working voltages (commonly above 4.0 V), low cost, and excellent safety of DIBs make them a worthy research field. 116,138 Among them, K-based DIB (PDIB) has emerged as a novel energy storage device within the last 3 years with the prosperity of PIBs. Like conventional PIBs, the electrochemical performance of PDIBs also relies on the potassium storage capability of anode material. In terms of existing anodes for PDIBs (such as K, Sn, and Pb), 139 there are often non-negligible problems concerning structural stability, capacity, reversibility, and safety. Thus, the development of high-performance anode materials is currently a hot topic. It is believed that the modified metal chalcogenides for PIBs can be easily transplanted to aid the research of PDIBs because they share a similar working principle and performance requirements in terms of the anodes that are responsible for reversible potassium storage.
In fact, this possibility has been partially supported by the successful application of metal chalcogenides (such as Ni 3 S 2 ,WS 2 , and MoS 2 ) as efficient and durable anodes for lithium (LDIBs) and sodium DIBs (SDIBs). For example, Bonaccorso's group verified the possibility of employing few-layer WS 2 flakes as lithium storage materials for LDIBs. 140 The WS 2 anodes were used in DIBs based on PF 6 − and Li + insertion and yielded a charge capacity of 457 mAh g −1 and a wide cell voltage window of 0-4 V. Jiao's group used a 3D skeleton nanostructured Ni 3 S 2 /Ni foam coated with rGO as a high-performance DIB anode. 141 The assembled LDIBs operated in a voltage window of 4.2 to 4.5 V and showed excellent long-term cycling stability over 500 cycles with 85.6% capacity retention. In addition, MoS 2 /N-doped carbon hollow spheres, 142 penne-like MoS 2 /carbon nanocomposites, 143 and layer-by-layer stacked N,S-doped carbon/MoS 2 nanosheets 144 have also been fabricated for SDIBs using an NaPF 6 electrolyte. All exhibit stable cycling behavior but, notably, the reversible capacities are a little lower than those of LDIBs due to the lower sodium storage ability of metal chalcogenides.
On the basis of the aforementioned advances in the use of metal chalcogenides for DIBs, we believe highperformance PDIBs with metal-chalcogenide-based cathodes will be established as long as matching electrodes with suitable electrolytes and additives can be identified.

| Cathode catalysts for potassiumoxygen batteries
Alkali metal-O 2 batteries using aprotic electrolytes have received tremendous interest over the past decade because of their very high theoretical energy densities (eg, 3500 Wh kg −1 for Li-O 2 batteries and 350 Wh kg −1 for LIBs). Although alkali metal ions are not stored in the inner bulk of the cathode catalyst, they provide catalytic sites for the electrochemical reduction/oxidation of discharge products at the surface. Therefore, the catalysts play a key role in regulating the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) activities of air cathodes, for which the components and microstructures directly determine the final capacity, overpotential, and lifespan of metal-oxygen batteries. Among the various categories of catalysts for Li-O 2 batteries, metal chalcogenides have already been utilized to boost the reversible formation of Li 2 O 2 on the cathode. For example, Yin's group utilized hierarchical NiCo 2 S 4 -based core-shell nanostructures as cathode catalysts for Li-O 2 batteries. 145 The specific capacity, discharge/charge reversibility, and cycling performance have all been improved. Kim's group employed NiS 2 nanosheets as cathode electrocatalysts, yielding a Li-O 2 battery with a lifetime of more than 300 cycles. 146 Other metal chalcogenide species have also proven to be beneficial for Li-O 2 batteries, such as ZnCo 2 S 4 , 147 CoNi 2 S 4 , 148 and CuCo 2 S 4 . 149 Notably, compared to research into Li-O 2 batteries, that into K-O 2 electrochemistry is currently in the early stages. In fact, the ORR and OER mechanisms are much different in K-O 2 batteries. 101,130 KO 2 is believed to be the final stable discharge product (instead of K 2 O 2 , the potassium analog of the Li 2 O 2 that is formed in Li-O 2 batteries). Some experimental results gave an evidence for this statement. For example, Wu's group proposed a K-O 2 battery based on O 2 − -transfer on a carbon cathode. 130 In this system, first, the oxygen reacts with K + ions in the electrolyte to form KO 2 crystals, and, second, electrons migrate through the surfaces of the KO 2 crystals and react with oxygen and K + ions, enabling the growth of KO 2 from small crystals to nanosized particles and films. However, KO 2 rapidly covered the surface of the carbon cathode during the discharge process, resulting in an increased kinetic barrier due to the depletion of active sites exposed to oxygen. Besides, high currents also induced larger oxygen gradients on the porous carbon cathode, resulting in the less-efficient loading of discharge products. These are the key factors causing K-O 2 battery failure. Superior catalysts are expected to address the difficulties plaguing existing K-O 2 batteries.
On account of their enhanced electrocatalytic effects in Li-O 2 batteries, we think that metal chalcogenides will also be favorable for K-O 2 batteries. This consideration is mainly based on the following viewpoints. First, it is recognized that superoxides have a higher oxidizing ability than hyperoxides, so carbon electrode corrosion is more serious in K-O 2 batteries, which limits cyclability. Metal chalcogenides are effective electron donors, and the unpaired electrons from dangling chalcogen sites can neutralize the oxidizability of KO 2 , reducing the structural damage and extending the longevity of the cathode catalysts. Second, the unoccupied 3d orbitals of transition metal atoms can act as mediators to accommodate the released electrons during the decomposition of discharge products and increase the fast transport of electrons from the electrode materials to the current collectors.
Generally, because these functions of metal chalcogenides have been demonstrated in the above aprotic Li-O 2 cases and other metal-O 2 batteries, where the cells with metal chalcogenide-based cathodes showed better electrochemical performance than their counterparts with pure carbon cathodes, [150][151][152] it is reasonable to further expect them to become cost-effective, efficient, and durable cathodes for K-O 2 batteries.

| SUMMARY AND PERSPECTIVE
Metal-chalcogenide-based composites exhibit great potential for potassium storage based on the findings of previous investigations. The depth of potassiation and depotassiation is strongly dependent on the reaction kinetics of metal chalcogenides, which is affected by their components, microstructure, particle size, and hybridization. Four dominant modification strategies including limitation of discharge depth, nanostructure engineering, confinement by carbon nanophases, and ternary alloying can be used to resolve the intrinsic structural instability and low electric/ionic conductivity of metal chalcogenides, which enable obvious improvements in electrochemical performance. However, these modification strategies still raise questions worth consideration. In the end, based on our thinking, the following points should receive more attention in future research focusing on metal chalcogenides for potassium storage.
1. Investigate the potassiation/depotassiation behavior of unrevealed metal chalcogenides. Despite several metal chalcogenides have already proven to be with electrochemical activity for PIBs, it cannot exclude the feasibility that those unexplored counterparts may demonstrate more superior potassium storage capability. 2. Construct more advanced metal-chalcogenide-based nanostructures or nanocomposites to facilitate longterm cycling stability at high rates. Multidimensional (such as 1D and 3D, 2D and 3D, and 0D and 3D) metal chalcogenides are even more promising than those limited to a single dimension because of the combination of their corresponding advantages. Other highly conductive and ultrastable substrates besides carbon nanophases can be used for the hybridization of metal chalcogenides to enhance their structural stability and reaction kinetics. 3. Construct metal chalcogenide-based heterojunctions. This direction seems solely reported for metal chalcogenidesbased PIB anodes. Nevertheless, surface/interface regulation should be an effective means that may direct the modifications to realize high performance. 4. Introduce more capacitance-controlled electrochemical behavior. At present, most of the reported metalchalcogenide-based PIB anodes have a limited capacitance-based capacity contribution (<80%), even at high rates. If this part of the capacity contribution can be promoted, the capacity retention at high rates should be enhanced. This would be especially useful for the fast charging required in consumer markets. 5. Devote more efforts into the electrolyte research studies. The suitable coupling of potassium salt and solvent is crucial to forming a stable and efficient SEI film on metal chalcogenides, which is a decisive factor affecting the capacity and cyclability of PIBs. To date, little attention has been paid to the electrolyte. 6. Perform more in situ characterization to understand the underlying potassium storage mechanisms. It is believed that the deeper and clearer understanding of mechanism will help the design of new metal-chalcogenide-based nanostructures, as well as provide inspiration for material modification in the future. 7. Fabricate novel potassium-based hybrid energy storage systems that combine the advantages of different energy storage devices. These new hybrid systems are expected to utilize the superior potassium storage capability of modified metal chalcogenides better, thus obtaining higher energy densities.
It is believed that the challenges facing current metal chalcogenides for PIBs and other potassium-based energy storage systems will be gradually overcome by advances in design, preparation, and characterization techniques, making potassium-based energy storage solutions a practical option for use in daily life. from Xinjiang University (China), respectively. Her current research interests mainly focused on the energy storage such as lithium ion batteries, sodium ion batteries, potassium ion batteries, and hydrogen storage, as well as electrochemistry characterization and computer modeling.