Framework materials for supercapacitors

2.1.1 Monometallic MOF

The limited electrical conductivity of pristine MOFs considerably hinders their application in supercapacitors [72,73]. High-performance electrode materials, especially conductive nickel-based MOFs, are outstanding candidates for design and synthesis to obtain unique pristine MOFs that can deliver great conductivity and cycling stability [74,75]. Sheberla et al. prepared conductive Ni-MOF Ni₃(HITP)₂ (HITP = 2,3,6,7,10,11-hexaiminotriphenylene) as electrode materials through the coordination interaction between Ni ions and HITP ligands without any conductive additives or other binders. This research demonstrated the special performance of conductive Ni-MOF with a favorable areal capacitance value of 18 µF cm⁻² at 0.05 A g⁻¹ and outstanding electrochemical stability, retaining 90% of the initial values after 10,000 cycles [61]. Novel one-dimensional (1D) Ni-MOF nanorods ([Ni(HOC₆H₄COO)_1.48(OH)_0.52·1.1H₂O]) with uniform morphology were successfully fabricated by Xu et al. under mild conditions [76]. In 6 M KOH, the Ni-MOF presented 1,698 F g⁻¹ at 1 A g⁻¹. Meanwhile, a promising cyclic retention of 94.8% was achieved over 1,000 cycles. Furthermore, in an assembled hybrid supercapacitor (HSC)—Ni-MOF//graphene, the HSC showed 166 F g⁻¹ at 1 A g⁻¹, exhibiting a promising capacitive property. Jiao et al. designed a special HSC that assembles Ni-MOF with CNTs-COOH in a solution with 3 M KOH and 0.1 M K₄[Fe(CN)₆], delivering 55.8 W h kg⁻¹ and a power density of 7,000 W kg⁻¹ [77]. These favorable performances can be ascribed to Ni-MOF that utilizes the redox reaction to store energy, exhibiting a large surface area, high porosity, and owe a unique 1D chain structure.

A hexagonal Ni-MOF (denoted as [Ni(HBTC)(4,4′-bipy)]) (HBTC = 1,3,5-benzenetricarboxylic acid, 4,4′-bipy = 4,4′-bipyridine) material was fabricated by Li et al. through a one-pot hydrothermal method [78]. They observed that the (001) crystal plane displays an outstanding structure for electron transport and electrolyte ion diffusion. In addition, the cross-inked mesh and porous nanostructure coating on the crystal plane (001) also facilitated the diffusion and storage of electrolyte ions compared with the crystal plane (100), indicating ideal electrochemical performance for supercapacitors (Figure 1a). From the scanning electron microscope (SEM) images, the Y3 sample displayed a clear mesh structure (Figure 1b). Hence, the electrode fabricated with Ni-MOF in the Y3 sample exhibited 977.04 F g⁻¹ at a current density of 0.5 A g⁻¹ as well as 92.34% capacitance retention over 5,000 cycles (Figure 1c). Galvanostatic charge/discharge (GCD) curves of Y1-5 samples at 0.5 A g⁻¹ indicated that the Y3 sample was superior to the other four, demonstrating perfect cycling stability. Moreover, the constructed asymmetric supercapacitor (ASC) manifested an outstanding energy density value of 55.26 W h kg⁻¹ with 362.50 W kg⁻¹ power density, while maintaining 33.98 W h kg⁻¹ with 3624.87 W kg⁻¹. Its excellent performances were ascribed to its layered framework structure, the (001) crystal plane, and cross-linked nanoporous nanostructure coating on the (001) crystal plane.

Figure 1

(a) Schematic illustration of charge transfer process on (001) and rectangular crystal planes, respectively. (b) SEM image of hexagonal Ni-MOF Y3 sample. (c) Cyclic performance of hexagonal Ni-MOF Y1-5 samples at 0.5 A g⁻¹. Reproduced with permission from [78]; Copyright 2019, Wiley. (d) The preparation of ultrathin 2D Co-MOF nanosheets. (e) Schematic illustration of the length of electron transportation in nanoparticle-based electrode. (f) SEM image of ultrathin 2D Co-MOF nanosheets. (g) Cycling performance of ultrathin 2D Co-MOF nanosheets at a current density of 2.0 A g⁻¹. Reproduced with permission from [83]; Copyright 2019, Elsevier. (h) Crystal structure of Cu–CAT viewed along the c-axis. (i) TEM image of Cu–CAT nanowire (insets: SAED pattern). (j) Comparison of the specific capacitances of the electrodes. (k) Cycling performance of Cu–CAT NWA-based supercapacitor. Reproduced with permission from [85]; Copyright 2017, Wiley. (l) Schematic illustration of the synthesis process of Cu-DBC. (m) HRTEM image taken along c-axis. (n) Energy density and power density of comparison with other reported materials. Reproduced with permission from [86]; Copyright 2019, Wiley.

Apart from Ni-MOF, Co-MOF is also of great interest owing to its conjugate π bonds, excellent electrolyte penetrability, controllable functionality, and lower steric hindrance, which makes it a good solution to conductivity and stability problems [79]. Sanati et al. fabricated a new Co-MOF (Co (II)-TMU-63, {[Co₃(μ4-tpa)₃ (μdapz) (DMF)₂]·2DMF} _n , H₂tpa = terephthalic acid, dapz = pyrazine-2,5-diamine), and an assembled Co(ii)-TMU-63//AC device was successfully built. It was tested in a 2 M KOH electrolyte; the specific capacitance of this electrode material was up to 384 F g⁻¹ at 6 A g⁻¹, while maintaining approximately 90% of the original values over 6,000 charge–discharge cycles [80]. Liu et al. obtained a layered Co-LMOF that exhibited a specific capacitance of 2,474 F g⁻¹ at 1 A g⁻¹ and 94.3% retention over 2,000 cycles [81]. These reports demonstrate that a high-performance specific capacitance may have a trade-off with high stability and vice versa. Typically, EDLC materials have long cyclic lives, high power, and low capacitance performances. Some frameworks with pseudocapacitive properties usually deliver higher capacitance values but a slightly shorter cycling life than their EDLC counterparts [82]. To obtain MOFs with both high-performance capacitance and high cycling stability, Zheng et al. prepared an ultrathin two-dimensional (2D) Co-MOF nanosheet (Co₂(OH)₂BDC, BDC = 1,4-benzenedicarboxylate) by controllable one-pot facile hydrothermal preparation methods (Figure 1d) [83]. The SEM image showed that the ultrathin Co-MOF had a more regular and larger layered structure than the bulk and micro–nano samples (Figure 1f), which could offer numerous redox-active sites for the rapid surface redox reaction and display desirable pseudocapacitive behaviors. Owing to the ultrathin structure of the 2D Co-MOF, its conductivity could be enhanced when mixed with carbon additives (Figure 1e). When fabricated with ultrathin 2D Co-MOF and tested in 3 M KOH, this electrode showed 1,159 F g⁻¹ at 0.5 A g⁻¹ and maintained at 96.7% during 6,000 cycles, where the electrode capacity changed from 1,106 to 1,070 F g⁻¹ at 2.0 A g⁻¹ (Figure 1g). These behaviors could be ascribed to the ion-exchange mechanism. Furthermore, an aqueous device with three different morphologies (ultrathin, micro–nano, and bulk) of Co-MOFs (positive) and active carbon (AC) (negative) was constructed. The ultrathin morphology delivered the best electrical performance in the ultrathin sample compared with the other two samples. Their research also proved that ultrathin 2D nanomaterials are promising novel nanomaterials owing to their excellent electrochemical properties [84]. The research in MOFs with unique ultrathin morphologies is reasonable.

Designing and synthesizing unique MOFs with other metal ions may also improve the conductivity of pristine MOFs. Conductive MOF Cu-CAT nanowire arrays (NWAs) were fabricated by Li et al. for the first time [85]. These NWAs were constructed via the coordination between Cu ions and ligands HHTP (HHTP = 2, 3, 6, 7, 10, 11-hexahydroxytriphenylene), forming honeycomb-like porous structures (Figure 1h). Numerous oriented and high-density nanowires homogeneously wholly covered the fibers and formed a uniform hexagonal top facet and hexagonal-prism shape. This could facilitate the transportation of electrolyte solutions to realize high-rate behaviors. HRTEM image and SAED revealed the single-crystal features of NWAs (Figure 1i). In the 3 M KCl solution, the electrode fabricated using Cu-CAT NWAs could exhibit 202 F g⁻¹ at 0.5 A g⁻¹ compared with other well-known low-conductivity MOF materials (Figure 1j). It also maintained 80% of the initial values after 5,000 cycles. Moreover, a solid-state supercapacitor based on Cu-CAT NWAs shows perfect stability, retaining more than 85% of its initial value after 5,000 cycles (Figure 1k). Liu et al. prepared a new conductive Cu-MOF (Cu-DBC) based on linker 8OH-DBC (8OH-DBC = dibenzo[g,p]chrysene-2,3,6,7,10,11,14,15-octaol) (Figure 1l) [86]. High-resolution transmission electron microscopy (HRTEM) image displayed a uniform microcrystalline structure, indicating its high crystallinity and perfect orbital interaction (Figure 1m). Owing to the perfect orbital interaction between Cu ions and 8OH-DBC, these Cu-MOFs could possess excellent chemical stability. Hence, Cu-DBC could be used in nonflammable and low-cost aqueous electrolytes compared with other traditional MOFs. In 1 M NaCl solution, its gravimetric capacitance calculated from the discharge curve was 479 F g⁻¹ at 0.2 A g⁻¹. The electrode materials also maintained 72% of their initial capacity during 2,000 cycles, even at 5 A g⁻¹. Moreover, the assembled symmetric solid-state supercapacitor based on Cu-DBC electrodes exhibited 13.8 W h kg⁻¹ energy density with 0.1 kW kg⁻¹ power density and maintained 7.6 W h kg⁻¹ under 5.0 kW kg⁻¹ (Figure 1n). This indicated that supercapacitors based on Cu-DBC frameworks possess promising energy density and power density as well as better electrochemical performance than other reported conductive MOFs.

2.1.2 Multimetallic MOF

The design and construction of monometallic MOFs with unique structures have received significant attention [87]. However, monometallic MOFs still have shortcomings, such as low specific capacitance values and unsatisfactory stability. Hence, the design and fabrication of multimetallic MOFs as electrode materials is an effective and outstanding strategy for addressing these problems [88]. Multimetallic MOFs, also denoted as mixed metal–organic frameworks, embedded various central metal ions into the same framework structure. This not only enhances the stability of the pristine MOF structure but also improves the framework stability and the tuning of breathing behavior, thereby delivering better performance compared to that of monometallic MOFs [89,90]. To develop multimetallic MOFs, two or more different metal ions can be incorporated into the same frame structure. This would provide MOFs with additional degrees of structural stability and endow them with new behaviors [60]. However, research on multimetallic MOFs in the field of supercapacitors remains largely unexplored, which may become one of the breakthroughs in improving the electrochemical performance of supercapacitors.

In the recent research, Ni/Co-MOFs have exhibited uniform size, oriented morphology, and ordered distribution, which could enhance their electrochemical performance [93]. To explore their applications, Zhang et al. synthesized a lamellae-like Ni/Co-MOF through a facile hydrothermal route (Figure 2a) [91]. With the combination of multimetallic components and special lamellae-like nanostructures, the fabricated Ni/Co-MOF_0.75 displayed a significant improvement in capacitance values and cycling stability. The morphology analysis indicates the ultrathin nanosheets, which are conducive to its electrochemical properties (Figure 2b). The specific capacitance of this electrode was 568.0 C g⁻¹ at 1.0 A g⁻¹ (Figure 2c), along with the 75.5% initial values remaining after 3,000 cycles. In an assembled ASC based on this electrode material with the reduced graphene oxide (rGO) (Figure 2d), this device delivered excellent cycling performance, with 82.6% capacitance retained during the 10,000 cycling tests. It also delivered 42.24 W h kg⁻¹ along with a power density of 800 W kg⁻¹. Bu and coworkers also prepared Ni/Co-MOF (Ni/Co-terephthalic acid/NF) nanosheet arrays on conductive substrates via a facile one-step synthesis method (Figure 2e) [92]. According to the SEM image, the 3-2 NiCo-MOF (Ni:Co = 3:2) displayed a triangle-like nanosheet morphology and completely covered the nickel foam surface (Figure 2f). This could produce a three-dimensional (3D) shape network with numerous open spaces, making more efficient and faster ion charge transfer and electron transport possible. When tested in 6 M KOH solution, the 3-2 NiCo-MOF sample preserves 1003.5 C g⁻¹ (2,230 F g⁻¹) at 1 A g⁻¹. In a hybrid supercapacitor (assembling Ni/Co MOF with AC), this device exhibited 34.3 W h kg⁻¹ energy density under 375 W kg⁻¹ power density and 20.83 W h kg⁻¹ energy density with 7.50 kW kg⁻¹ power density values (Figure 2g). Remarkably, this HSC maintained approximately 75.2% of the initial capacitance values during 6,000 cycles.

Figure 2

(a) Schematic illustration of the synthetic process of the Ni/Co-MOFs. (b) SEM image of Ni/Co-MOF_0.75 sample. (c) Comparison of specific capacitances of MOFs. (d) Electrode design of the present ASC. Reproduced with permission from [91]; Copyright 2020, Elsevier. (e) Schematic illustration of the formation of 3-2 NiCo-MOF nanosheets arrays on nickel foam. (f) SEM image of 3-2 NiCo-MOF/NF. (g) Ragone graphs of the 3-2 NiCo-MOF/NF//AC device. Reproduced with permission from [92]; Copyright 2019, Elsevier.

Remarkable works on pristine MOF as electrode materials for supercapacitors are listed in Table 1.

2.2.1 Porous carbon

Nanoporous carbon is an excellent supercapacitor candidate owing to its optimized pore size distributions, perfect thermal and physiochemical stability, high surface area, and low cost for regeneration [94,95]. These advantages offer a shorter pathway for the diffusion of electrolyte solutions, buffering the volume charge during the cycling process and obtaining more electrochemically active sites for the adsorption and desorption of electrolyte ions [96]. Hence, designing and synthesizing nanoporous carbon materials as electrode materials for supercapacitors is necessary. In the previous research, MOFs possessing high SSAs and highly customizable pore structures have been explored as perfect templates for preparing carbon materials to support capacitive active materials [97,98]. MOF-derived carbon materials have attracted significant attention owing to their controllable morphologies and pore structures [99]. The porous carbon derived from MOFs can possess abundant nanoscale cavities and provide larger surfaces for electron desorption–adsorption as electrode materials for supercapacitors [100].

To further study MOF-derived porous carbon materials, Qian et al. chose a zinc-organic framework as a precursor to prepare robust cage-based mesoporous carbon materials (BMM-9) through carbonization at high temperatures [101]. BMM-9-900 (calcinated at 900°C) exhibited 182.8 F g⁻¹ at 1.0 A g⁻¹, and the stability texts showed that BMM-9-900 retained 98.5% initial values at 10 A g⁻¹ during 1,000 cycles. Khan et al. utilized MOF-5/AC as a soft template to derive carbon nanospheres via carbonization at 850°C, exhibiting 300 F g⁻¹ at 1.5 A g⁻¹ along with 91.5% capacitance values retained during 3,000 charge–discharge cycles [102]. Recently, some researchers observed that nitrogen doping in carbon materials can improve the surface polarity and humidity, ultimately optimizing the electrical conductivity, which could also improve specific capacitance [103,104,105]. Tang et al. prepared flexible 3D N-doped porous carbon bubbles (CTAs@NCBs) connected via carbon tube arrays derived from a ZnO@ZIF-8 nanoarray template [106]. This material-assembled supercapacitor yielded an areal capacitance of 580 mF cm⁻² in 1 mA cm⁻², and 98.5% of its initial capacitance remained over 10,000 cycles. This could be ascribed to its structural characteristics because it possessed excellent ionic conductivity and a high surface area.

Although porous carbon derived from MOFs has attracted increasing attention, partial breakdown in the initial MOF structure during the carbonization and annealing processes cannot be completely avoided. This results in a decrease in the surface area and redox-active sites, limiting the conductivity of the as-prepared nanoporous carbonaceous materials and their rate performance [108]. To solve the problems, Wang et al. fabricated a unique porous carbon possessing a “brick-and-mortar” sandwiched structure by using MOF-5-derived nanoporous carbon film as “mortar” and the graphene nanosheets as “brick” (Figure 3a) [107]. This porous carbon material (denoted as C-GMOF) remained a layered structure, and TEM analysis confirmed the formation of this morphology (Figure 3b). A sandwich-shaped structure with a superior surface area was conducive for convenient electron transportation with less transport length and charge storage. In an electrochemical test, the GCD curve was almost linear and symmetrical, indicating its excellent reversibility (Figure 3c). This C-GMOF-based electrode material exhibited 345 F g⁻¹ at 2 mV s⁻¹, along with 201 F g⁻¹ at 2 V s⁻¹, which was relatively higher than other materials (Figure 3d). Furthermore, the specific capacitance remained at 99% of its initial values after 10,000 cycles (Figure 3e), suggesting its outstanding cycling stability.

Figure 3

(a) Schematic illustration of the fabrication process for porous carbon building using MOFs and GO as precursors. (b) TEM image of the C-GMOF. (c) GCD curves of the C-GMOF, C-MOF, YP-50, and RGO at 50 A g⁻¹. (d) Specific capacitances of the C-GMOF, C-MOF, YP-50, and RGO at various scan rates. (e) Cycling performance of the C-GMOF at 200 mV s⁻¹, and the inset shows CV curves before and after 10,000 cycles. Reproduced with permission from [107]; Copyright 2016, Elsevier.

2.2.2 Metal oxides

Metal oxides, especially transition metal oxides (TMOs), have attracted considerable attention in the scientific community as electrode materials in supercapacitors owing to their high electrochemical response, perfect electrical conductivity, low manufacturing costs, large SSA, and easy processability [109,110]. TMOs such as Co₃O₄, NiO, RuO₂, Mn₂O₃, Fe₂O₃, and some binary transition metal oxides have already been extensively researched for application in supercapacitors [111]. Among these TMOs, Co₃O₄, whose theoretical specific capacitance can reach up to 3,560 F g⁻¹, is one of the most widely utilized oxides in electrode materials for supercapacitors [112,113]. Previous research has demonstrated that Co₃O₄ can be obtained via the calcination of ZIF-67 precursors [114]. However, approaching the theoretical capacitance of Co₃O₄ is still difficult, which limits its application in supercapacitors considerably [115]. To improve their capacity and cycling performance, Li et al. prepared nanoporous Co₃O₄ nanotubes by calcining nanoscale Co-MOF-74 crystals at 350°C [116]. A hexagonal cuboid morphology was achieved during the annealing process. In the electrochemical tests with 2 M KOH electrolyte, Co₃O₄-350 presented 647 F g⁻¹ at 1 A g⁻¹, and there was no obvious decrease during the 1,500 cycles tests in 2 A g⁻¹. Liu et al. fabricated dodecahedral NaBH₄-reduced Co₃O₄ (R-Co₃O₄) through a conformal transformation route [117]. In their work, NaBH₄ was used to modify the surface state of Co₃O₄ that displayed a dodecahedral morphology, uniform dimensions, and rhombic dodecahedron morphology. This resulted in a capacitance of 329 F g⁻¹ at 1 A g⁻¹, and 92% of the original values were retained during 1,000 cycles.

Although Co₃O₄ has emerged as an excellent electrode material, researchers still observed that it has limitations such as low ionic diffusivity, lack of outstanding electron conductive routes, and drastic volume expansion [75]. The incorporation of conductive substrates such as graphene or carbon cloth is a promising solution to these problems [118]. Vilian et al. incorporated rGO onto MOF-derived Co₃O₄ via a hydrothermal route [119]. In this work, Co₃O₄ has a hexagonal sheet-like structure, which assists in the electron accessibility of the electrode materials. The rGO-Co₃O₄ delivered 1,300 F g⁻¹ and 80.5% of its original capacitance over 5,000 cycles in 4 M KOH at 4 A g⁻¹. Furthermore, the assembled ASC exhibited 65.8 W h kg⁻¹ energy density values under a 2.05 kW kg⁻¹ power density. Yin et al. prepared Co-MOF-derived Co₃O₄ coated with the NiCo layer double hydroxide (NiCo-LDH) nanosheets on a carbon cloth [120]. They first annealed Co-MOF on a carbon cloth to fabricate derived cobalt oxides and then combined them with NiCo-LDHs to construct hybrid Co₃O₄@NiCoLDH. As calculated from the electrochemical tests, this hybrid showed 1,708 F g⁻¹ at 1.0 A g⁻¹, representing 89% of the original capacitance values retained during 4,000 cycles at 10 A g⁻¹. Furthermore, the constructed ASC (assembled with AC) delivered 44.44 W h kg⁻¹ energy density under 0.80 kW kg⁻¹ power density.

Apart from Co₃O₄, NiO is also a competitive candidate for electrode materials in supercapacitors owing to its rich oxidation states, abundant SSA, low cost, thermal stability, and pronounced electrical properties [121]. NiO delivers perfect capacitive properties, expressing 2,584 F g⁻¹ theoretical specific capacitance values [122,123]. Wu et al. fabricated Ni-MOF-derived NiO nanospheres via a simple solvothermal method [124]. After calcination at 400°C, NiO hollow nanospheres with highly uniform shapes were successfully obtained (denoted as N400) with a core-in-double-shell structure. This N400 sample exhibited promising electrochemical performance with 473 F g⁻¹ at 0.5 A g⁻¹, and 94% original values were retained during 3,000 cycles. In an asymmetric supercapacitor based on N400 and AC (N400//AC), this device exhibited 12.5 W h kg⁻¹ energy density along with a power density of 7.5 kW kg⁻¹. Fabricating a hybrid nanocomposite is an effective strategy to further improve their electrical conductivity and obtain the theoretical specific capacitance. Li et al. prepared nanorod-composed yolk-shell microspheres derived from Ni/Zn-MOF via a simple solvothermal method [125]. In this study, the specific capacitances of the material calculated were 497 F g⁻¹ at 1.3 A g⁻¹, along with almost 100% of the original values retained during 2,000 cycles, indicating its excellent cycling performance.

Other monometallic oxides have also been developed for use in electrode materials. Yuan et al. fabricated layered δ-MnO₂ nanostructure materials using [Mn(C₈H₄O₄)·(H₂O)₂] _n as a precursor to react with KMnO₄ via a hydrothermal method [126]. Its specific capacitance can deliver 416 F g⁻¹ at 0.5 A g⁻¹. The assembled ASC (MnO₂//AC) exhibited 23.2 W h kg⁻¹ with a power density of 425 W kg⁻¹. Owing to its layered nanostructure, access to cations from the electrolyte could be accelerated, thus facilitating promising electrochemical performance. Wang et al. synthesized ZnO/C materials by directly transforming MOFs into hollow carbonaceous materials and then reacting them with glucose-derived polymers to form hollow Zn-coated carbonaceous composites [127]. ZnO/C-based electrode materials can deliver 394 F g⁻¹ at 1 A g⁻¹, indicating its promising electrochemical performance.

Recently, some binary or mixed metal oxides derived from multimetallic MOFs have been widely investigated owing to their high SSA, unique nanostructure, and richer active sites [128]. In addition, their uniform metal distribution is beneficial for reducing lattice distortion and further enhancing the capacity and cycling stability [129]. MOF-derived mixed metal oxides display noticeable strengths compared with their single pure metal oxide counterparts [130]. Zhang et al. utilized an Fe-MOF gel (MIL-100-Fe) as a template to fabricate mesoporous LaFeO₃ perovskite nanoparticles [131]. Owing to its mesoporous structure and 2 nm concentrated pore size, this synthesized LaFeO₃ showed 241.3 F g⁻¹ at 1 A g⁻¹. Furthermore, the all-solid-state symmetric supercapacitor based on LaFeO₃ delivered 34 W h kg⁻¹ energy density values with 900 W kg⁻¹ power density, along with the 92.2% initial values remaining after 5,000 cycles. Functional additives can be incorporated into these binary or mixed metal oxides to further improve their electrochemical performance. Xie et al. introduced MXene nanosheets into derived CoFe₂O₄ nanorods [132]. This could facilitate charge transfer, maintain its excellent flexibility, and enhance its ion transmission path. In the 1 M LiCl electrolyte, the CoFe₂O₄/MXene composite exhibited an excellent volumetric capacity of 2467.6 F cm⁻³. Moreover, a flexible symmetrical supercapacitor retained 88.2% original values over 10,000 cycles. This indicated its promising application potential in supercapacitors. Yin et al. synthesized hollow NiCo₂O₄ nanocages grown on a SiC nanowire network/carbon cloth [133]. With unique structural advantages such as a hierarchical porous structure and interwoven conductive networks, this material exhibited a specific capacitance of 1,377 F g⁻¹ at 1 A g⁻¹ and 88.3% capacitance retention over 6,000 cycles. When assembled with AC as a supercapacitor, this supercapacitor delivered 46.58 W h kg⁻¹ energy density with 0.80 kW kg⁻¹ power density.

Among the diverse binary metal oxides derived from bimetallic MOFs, nickel–cobalt binary oxides have attracted considerable interest from the scientific community because of their high-power supercapacitor applications. Yu et al. prepared a NiO/NiCo₂O₄ (1:1) derived from MOF-74-Ni/Co MOF (1:1) (Figure 4a) [134]. The morphology analysis showed that the mixed metal oxide samples all inherited the morphology of their precursors with nanowire shapes (Figure 4b). At different Ni/Co ratios, they observed that the NiO/NiCo₂O₄ (1:1) sample displayed a hollow nanowire structure, indicating its porous nature. In this study, the calculated 732.0 C g⁻¹ derived from NiO/NiCo₂O₄ (1:1) was much higher than of the other samples and monometallic oxides at 1 A g⁻¹ (Figure 4c). It also retained 92.5% initial capacitance values at 10 A g⁻¹ over 3,000 cycles, proving its excellent stability (Figure 4c). In an asymmetric supercapacitor (NiO/NiCo₂O₄ (1:1)//AC), the specific energy density reached 36.6 W h kg⁻¹ along with a 6791.8 W kg⁻¹ power density, which was much better than the reported values (Figure 6c). Jayakumar et al. fabricated Ni _x Co_3−x O₄ derived from ZIF-67 and then synthesized graphene/NiCo oxide hybrids (GNi:Co) [135]. During the hydrothermal treatment, the GNi:Co (Ni:Co = 1:1) composite hydrogel displayed a highly microporous network with interlinked graphene layers on nanosized Ni/Co particles (Ni:Co ratio of 1:1). The morphology analysis showed that these samples possessed a hollow structure, thus helping to connect the graphene layers (Figure 4d). Owing to the conducive porous network from the surrounding graphene, this material showed better electrochemical performance than other compared counterparts. This GNi:Co 1:1 hydrogel displayed the highest specific capacitances of 2,870 F g⁻¹ at 1 A g⁻¹ compared with other samples (Figure 4e). Moreover, this material presented a significant enhancement in cycle life, maintaining 81% capacitance retention after 5,000 long cycles (Figure 4f). In an ASC assembled with pure graphene hydrogel, the ASC could provide a higher specific energy density of 50.2 W h kg⁻¹ along with 750 kW kg⁻¹ power density (Figure 4g). These comparisons could be useful in identifying this material for appropriate real-life applications.

Figure 4

(a) The synthesis process of metal oxides with hollow nanowires. (b) TEM image of an individual NiO/NiCo2O4(1:1) hollow nanowire. (c) Cycling stability at 10 A g⁻¹ for 3,000 cycles. Reproduced with permission [134]; Copyright 2018, The Royal Society of Chemistry. (d) SEM image of GNi:Co 1:1. (e) Specific capacitances of samples. (f) Cyclic performances of samples. (g) Ragone Plot of GNi:Co 1:1//Pure graphene with other reported works. Reproduced with permission from [135]; Copyright 2017, Wiley.

Figure 5

(a) Schematic illustration of MOF-74 converting to Ni–Co MDH. (b) SEM image of 65Ni-MDH. (c) Specific capacity of MDH electrodes with different initial Ni(ii)-to-Co(ii) ratios. (d) Cycling stability and faradaic efficiency of the 65Ni- MDH//N–C device at 13.5 A g⁻¹ for 10,000 cycles. Reproduced with permission from [140]; Copyright 2017, American Chemical Society. (e) Schematic illustration of the synthetic strategy of the Ni/Co-LDH. (f) Specific capacitance of Ni/Co-LDH electrodes at current densities from 1 to 25 A g⁻¹. (g) Cycling performance of the Ni/Co-LDH electrodes at a current density of 5 A g⁻¹. Reproduced with permission from [141]; Copyright 2019, American Chemical Society.

2.2.3 Metal hydroxides

In recent years, transition metal hydroxides have been considered as perfect electrode materials owing to their high surface area and ultrahigh theoretical capacitance (3,500–4,600 F g⁻¹) [136]. MOF-derived metal hydroxides are pseudocapacitive materials that can store numerous energies depending on the fast redox reactions that drastically improve energy storage and exhibit much higher discharge capacities than EDLC materials [137,138]. Wu et al. synthesized Co/Ni-double hydroxides with a novel wisteria flower-like microstructure on carbon cloth in the NiSO₄ solution [139]. This hybrid delivers 929.4 C g⁻¹ at 2 mA cm⁻². The assembled ASC exhibits 42.5 W h kg⁻¹ with 81.5% capacity remaining after 7,500 cycles. Qu et al. utilized a classic nickel–cobalt MOF-74 as a precursor to fabricate highly porous double hydroxide (MOF-74-derived double hydroxide, denoted as MDH) via a facile alkaline treatment (Figure 5a) [140]. The SEM image showed the formation of nanoscale 65Ni-MDH after KOH treatment of the as-synthesized MOF-74 templates under controlled hydrothermal conditions (Figure 5b). The SSA of MDH reached 299 m² g⁻¹. Its superior surface area could be ascribed to its inheritance from the porous nature of pristine MOFs, and the 65Ni-MDH displays promising porous properties. In the electrochemical test, this electrode material displayed 875 C g⁻¹ at 1 A g⁻¹, which was better than that of other samples (Figure 5c). The assembled HSC using 65Ni-MDH and N-C as electrodes could retain 91.3% after 10,000 cycles at 13.5 A g⁻¹, and the Faradaic efficiency was close to 100% during the cycling process (Figure 5d). It also showed 81 W h kg⁻¹ energy density along with 1.9 kW kg⁻¹ power density and maintained 42 W h kg⁻¹ under 11.5 kW kg⁻¹.

Figure 6

(a) Schematic illustration of the synthesis procedure of the R-NiS/rGO from MOF precursors. (b) Specific capacities of R-NiS/rGO based on different current densities. (c) Cycling performances of R-NiS/rGO and N-NiS. Reproduced with permission from [38]; Copyright 2018, The Royal Society of Chemistry. (d) SEM image of NiCo-S-1 samples. (e) Cycling performance of the NiCo-S-1//AC at 2 A g⁻¹. (f) Ragone plots of the current work and the other related devices. Reproduced with permission from [153]; Copyright 2020, The Royal Society of Chemistry.

Many researchers have recently explored MOF-derived LDH. These derivatives exhibit multimetallic hollow structures, enhancing the flow efficiency of ions and reducing the resistance of ion diffusion [142]. Among various LDH materials, Ni/Co LDH exhibits attractive performance, including high theoretical capacitance, rich redox reaction sites, and excellent electronic conductivity [143,144]. However, its easily agglomerated structure greatly hinders its further applications [145,146]. A promising strategy for overcoming this problem involves adjusting the morphology by fabricating a hierarchical structure for instance [147]. Xiao et al. utilized a Ni/Co-MOF (with Ni:Co = 7:3) to synthesize LDH with hierarchical microspheres (Figure 5e) [141]. In their work, they constructed hierarchical LDH microsphere structures via a controllable alkaline hydrolysis strategy. Owing to the high tunability of MOFs, Ni/Co LDH can have a desirable composition and morphology. Hence, the surface area of the Ni/Co LDH was measured to be 201 m² g⁻¹, making it applicable for supercapacitors. In the 1 M KOH solution, the Ni/Co LDH-based electrode material showed 1,652 F g⁻¹ at 1 A g⁻¹, presenting almost 100% value after 2,000 cycles. Both these properties are much better than those of other samples (Figure 5f and g). Moreover, an ASC composed of this LDH and AC delivered 32.9 W h kg⁻¹/74.3 W kg⁻¹ energy density/power density and maintained 26.0 W h kg⁻¹ under 1817.3 W kg⁻¹.

2.2.4 Metal sulfides

Transition metal sulfides have also become one of the most promising candidates for electrode materials in supercapacitors owing to their higher specific capacitances, higher electrical conductivity, and richer redox reactions than their oxide or hydroxide counterparts [39,148]. Some metal sulfides such as nickel sulfides and cobalt sulfides have attracted considerable attraction as promising candidates of high-performance supercapacitors [149]. However, it is difficult to deliver ideal capacity and cycling stability via using single-component forms of those sulfides [150]. To deal with these problems, strategies such as combining carbon-based materials (graphene, carbon nanotubes (CNTs), etc.) and other pseudocapacitive materials can improve their conductivity and enhance their cycling stabilities [151,152].

Qu et al. reported NiS nanorods combined with rGO derived from a Ni-MOF-74/rGO hybrid (Figure 6a) [38]. In this study, R-NiS/rGO exhibited a remarkable integral area and enriched active sites, implying a significantly improved specific capacitance. The NiS nanorod surfaces in this nanohybrid are very active for rapid redox reactions, indicating their promising pseudocapacitive properties. Moreover, the presence of rGO greatly enhanced its electronic conductivity, providing more efficient current collection for rapid energy storage. In the electrochemical test with 2 M KOH electrolyte, it demonstrated 744 C g⁻¹ in 1 A g⁻¹ and an impressive capacity retention of 89% over 20,000 cycles (Figure 6b and c). In HSC assembled with R-NiS/rGO and ZIF-8-modified nitrogen-doped graphene aerogel, 54 W h kg⁻¹/46,034 W kg⁻¹ energy density/power density was achieved.

Apart from monometallic sulfides, some bimetallic transition metal sulfides also possess better electrochemical activity, more varied stoichiometric chemistry, and richer redox reactions than single-component metal sulfides [154,155]. It was observed that bimetallic metal sulfides express lower kinetic energy barriers, which makes their charge transportation more effective than that in monometallic metal sulfides [156]. Zhao et al. fabricated porous NiCo₂S₄ nanosheet arrays with enriched edge sites on electrochemically activated flexible carbon cloth derived from Co-MOF for the first time [157]. The morphology analysis showed a dense and uniform coverage of the Ni–Co–S nanosheets on the carbon fiber surface. This leads to the formation of the desired hollow and porous structure, thus facilitating the diffusion of electrolyte ions and the insertion and disembedding process in the cycling stability test. This material-based electrode delivered 2,392 F g⁻¹ at 1 A g⁻¹ and a desirable stability routine with 85.1% capacity retention over 10,000 cycles at 30 A g⁻¹. Furthermore, in an assembled ASC, it exhibited 30.1 W h kg⁻¹/800.2 W kg⁻¹ energy density/power density. Zheng et al. developed ultrathin NiCo sulfide nanosheet arrays on Ni foam with robust adhesion [153]. Owing to the presence of Ni₃S₂ and Co(OH)₂ in the synthesis process, these materials could deliver ultrahigh capacity. The morphology analysis of NiCo-S-1 indicated that its unique structure offers abundant SSA and open channels for electrolyte transportation and penetration (Figure 6d). These properties can reduce the diffusion distance and enhance the electrochemical performance. With its superior structure, this material demonstrated the highest electrochemical performance, delivering an ultrahigh value of 3,724 F g⁻¹ at 1 A g⁻¹ (Figure 6e). In an assembled NiCo–S//AC supercapacitor, the capacitance retention was stable with 90% retention after 3,000 cycles. The ASC device also showed 44.76 W h kg⁻¹/0.80 kW kg⁻¹ energy density/power density and maintained 24.99 W h kg⁻¹ at 12.00 kW kg⁻¹ (Figure 6f). These results indicate that NiCo–S has excellent potential for application in high-performance supercapacitors.

2.2.5 Other metal compounds

Some metal compounds such as metal selenides, metal phosphides, and nitrile have attracted considerable attention. Transition metal selenides with promising electrochemical activity, promising electronic conductivity, and stability have been investigated for supercapacitor applications [158,159]. Chen et al. synthesized porous CoSe₂ nanosheets derived from MOFs [160]. Owing to the porous nanostructure of CoSe₂, it could express abundant contact interfaces between the electrode and electrolyte, reducing the diffusion path of ions for this fast electrochemical kinetic process. This device showed 713.9 F g⁻¹ in 1 mA cm⁻² accompanying the impressive cycling property that maintained 92.4% initial values in 5 mA cm⁻² after 5,000 cycles.

Sun et al. fabricated bimetallic copper cobalt selenide nanoparticles embedded in N-doped carbon layers (denoted as (CuCo)Se/NC) via using CuCo-ZIF as a precursor [161]. Their route involved the fabrication of CuCo-ZIF and the selenization of the as-prepared bimetallic MOF. In the electrochemical tests with 2 M KOH, this hybrid exhibited 121.4 C g⁻¹ at 1 A g⁻¹. This is ascribed to the incorporation of Cu elements in CoSe matrix and the combination of N-doped carbon layer. Furthermore, the constructed HSC device based on this hybrid delivered an energy density of 16.3 W h kg⁻¹ along with a power density of 155.3 W kg⁻¹. It also possessed an excellent cyclic stability, retaining 96% initial capacity over 5,000 cycles.

Transition metal phosphides have also attracted increasing attention owing to their metalloid nature and perfect electrical conductivity [162,163]. Chu et al. fabricated ZnCo phosphide (Zn_0.33Co_0.67P) on Ni foam with a special structure of ZnCo phosphide array [164]. This phosphide electrode delivered 2,115 F g⁻¹ at 1.0 A g⁻¹ with 83.0% initial values over 7,000 cycles, which is caused by the lower electronegativity of phosphorus. Furthermore, they assembled Zn_0.33Co_0.67P (positive) with Bi₂O₃ (negative) to form an alkaline supercapacitor. This device delivered 83.05 W h kg⁻¹ with a power density of 775.02 W kg⁻¹ and a superior power density value of 15642.58 W kg⁻¹ along with a 26.94 W h kg⁻¹ energy density.

Kshetri et al. fabricated MOF-derived cobalt telluride-carbon nanoporous structure composite on nickel foam [165]. The morphology analysis indicated a highly porous nature and a uniform distribution of the constituent elements. This could lead to the unique charge storage process and better electrochemical performances. In this work, the hybrid electrodes delivered an areal capacitance of 307.5 mF cm⁻² at 20 mA cm⁻². Moreover, the assembled symmetric supercapacitor delivered an energy density of 43.84 W h kg⁻¹ and a power density 738.88 W kg⁻¹. It also retains 21.95 W h kg⁻¹ energy density at a high power density of 6173.44 W kg⁻¹.

Liu et al. prepared Ni-doped Co–Co₂N with a heterostructure via facial thermal annealing of MOF-derived NiCo₂O₄ under an ammonia atmosphere [166]. At different nitridation temperatures, they observed that the sample nitridized in ammonia at 350°C exhibited the best morphology. This made its nanoflakes uniform in morphology, and they grew vertically on the carbon fiber surface, which offered large surface areas for electrolyte ion access. Moreover, its thickness (less than 10 nm) also allows for faster electron transport, enhancing its electrical conductivity. In 1 M KOH, the Ni-doped Co–Co₂N (also denoted as Ni/Co–N-350) delivered 361.93 C g⁻¹ at 2 mA cm⁻², which is much higher than that of other oxide counterparts. Remarkably, an ASC with Ni-doped Co–Co₂N//porous carbon was constructed, which maintained 82.4% of its capacity after 5,000 cycles and achieved 20.40 W h kg⁻¹ energy density with 9.85 kW kg⁻¹ power density. All of its properties indicate that Ni-doped Co–Co₂N is an advanced electrode material.

Remarkable works on MOF derivatives as electrode materials for supercapacitors are listed in Table 2.

2.3.1 Carbonaceous material

Carbonaceous materials, including graphene and its derivatives, CNTs, AC, and mesoporous carbon, possess a high surface area and provide EDLC [169,170]. An effective strategy to address the poor conductivity of MOFs is to combine them with these carbon materials [171]. The as-prepared hybrids with carbon-based materials can also facilitate rapid ion diffusion as well as the increased charge storage properties via Faradaic reactions, thereby combining physical and chemical charge storage mechanisms in single-electrode materials [168].

Graphene oxide and reduced graphene oxide (GO and rGO) are considered as outstanding electrode materials for supercapacitors [172,173,174]. Owing to their unique layered structure, their ion diffusion and transportation can be further enhanced. This could ensure excellent performances in charge storage and delivery [175]. Liu et al. prepared an Fe-MOF/graphene aerogel composite. It possessed a high charge–discharge rate, promising capacitive volume, and reliable cycling stability, exhibiting 353 F g⁻¹ at 20 A g⁻¹ while maintaining 74.4% of the original values retained after 10,000 charge–discharge cycles [176]. RGO has emerged as a unique additive that is incorporated into MOFs as electrode materials for SCs [177]. Gupta et al. fabricated a Cu-MOF/rGO composite via two steps [178]. They first synthesized Cu-MOF {[Cu₂(L)(H₂O)₂]·(5DMF)·(4H₂O)} _n (Figure 7a) through a solvothermal method. The Cu-MOF/rGO composites were then constructed via a facile ultrasonication approach. Owing to the synergism between Cu-MOF and rGO, this composite delivered high mechanical strength and conductivity (Figure 7b). The morphology analysis showed that its crease-type structures reduce its ion-diffusion path, accelerating the transport and diffusion of ions on the electrode surface (Figure 7c). The wavy nature of the rGO sheets also improved their cycling efficiency. Combining the porous nature of Cu-MOF and conducting nature of rGO, GCD analysis indicated that this composite (1-GCE) delivered a promising discharge time at 0.8 A g⁻¹. In the electrochemical test, the specific capacitance was calculated to be 462 F g⁻¹ at 0.8 A g⁻¹, with 93.75% of the original capacitance retained after 1,000 cycles (Figure 7d and e). This highlights that the Cu-MOF/rGO composite is a promising candidate for supercapacitors.

Figure 7

(a) Crystal structure of CuMOF framework. (b) The proposed approach for the construction of composite 1. (c) SEM image of CuMOF/rGO composites. (d) Specific capacitance as a function of current densities for CuMOF–GCE, rGO–GCE, and 1–GCE. (e) Cycling stability of rGO–GCE and 1–GCE over 1,000 cycles. Reproduced with permission from [178]; Copyright 2019, American Chemical Society.

Apart from graphene, other carbon materials such as CNTs, mesoporous carbon, and carbon nanofibers also possess good availability, low weight, high conductivity, and high chemical stability and are used as conductive materials for combining with MOFs. Rui et al. fabricated a multiwall CNT and ZIF-67-Co hybrid fiber via a wearable and flexible fibrous device [179]. In the electrochemical test, the ZIF-67-Co/CNTs-derived Co₃O₄/CNTs were 110 F cm⁻³. Kim et al. synthesized a composite denoted as Ni-MOF@mC (mC: mesoporous carbon) [180]. This composite exhibited 109 F g⁻¹ and 91% initial capacitance retention over 5,000 cycles. In a symmetric supercapacitor, the device presented 38.8 W h kg⁻¹ energy density and 21,005 W kg⁻¹ power density.

2.3.2 Metal and metal compounds

Although some carbonaceous composites can enhance the conductivity and overall electrochemical performance of MOF-based materials to a certain extent, they are usually affected by hidden flaws: (1) the limited contact areas between pristine MOF microparticles and conductive additives, (2) MOFs in impeding their aggregation, (3) blocking of diffusion pores due to the conductive additives on pristine MOF surfaces. This has a negative effect on the diffusion of ions and their accommodation in pristine MOFs, leading to undesirable electrochemical performance [181]. To realize higher specific capacitance values and excellent energy density and power density in supercapacitors, selecting proper electrode materials is essential [182]. Some metal compounds can serve as pseudocapacitive positive electrodes for storing charges through Faradaic redox reactions [183,184]. Designing and fabricating well-aligned transitional metal oxides@MOF composites is a good strategy for dealing with these problems [185].

Accordingly, Zhang et al. combined Ni-MOF with NiS₂ to fabricate a Ni-MOF@ NiS₂ nanocomposite [186]. According to the morphology analysis, this composite exhibited nanosheet arrays and possessed a unique heterostructure, which enhanced its conductivity and ability to hold the porous MOF pristine structures. Electrochemical tests revealed 1,128 F g⁻¹ at 2 A g⁻¹ for this composite-based electrode. Moreover, the assembled ASC delivered 23.5 W h kg⁻¹ energy density under 1.4 kW kg⁻¹ power density. This ASC also maintained 95.2% of its original capacity retained over 10,000 cycles at 5 A g⁻¹, proving its excellent application in supercapacitors. Yue et al. synthesized a hierarchical MoS₂@Ni-MOF composite structure [187]. The morphology of MoS₂@Ni-MOF in the SEM image indicated its perfect crystallinity and noticeable topographic contrast (Figure 8a). In 3 M KOH, the composite delivered 1590.24 F g⁻¹ at 1 A g⁻¹, which was much better than that of pristine MOFs and common MoS₂ (Figure 8b). Moreover, the cycling test also demonstrated the excellent stability of the composites, maintaining a retention of 87.97% after 20,000 cycles (Figure 8c). In an ASC assembled with AC, this supercapacitor displayed 72.93 W h kg⁻¹ and a power density of 375 W kg⁻¹, which was superior to that of the previously reported supercapacitors.

Figure 8

(a) SEM image of MoS₂@Ni-MOF. (b) Specific capacity at various current densities. (c) Cycling stability at a current density of 20 A g⁻¹. Reproduced with permission from [187]; Copyright 2019, Elsevier. (d) The preparation of ultrathin Cu-MOF@δ-MnO₂ nanosheets. (e) The specific capacitance of ultrathin Cu-MOF@δ-MnO₂ nanosheets, ultrathin δ-MnO₂ nanosheets, and Cu-MOF nanoparticles electrodes at different current densities. (f) Cycling stability of ultrathin Cu-MOF@δ-MnO₂ nanosheets, ultrathin δ-MnO₂ nanosheets, and Cu-MOF nanoparticles electrodes at 4 A g⁻¹. Reproduced with permission from [188]; Copyright 2018, The Royal Society of Chemistry.

Other transitional metal compounds, including nickel-based compounds and manganese compounds, possess the high electrochemical activity and good electrical conductivity, which are effective in improving supercapacitors [189,190]. Zhu et al. prepared a Co-MOF with a hierarchical structure on nickel foam (Co-MOF/NF) [191]. This composite presented 13.6 F cm⁻² at 2 mA cm⁻². The assembled ASC could achieve 1.7 mW h cm⁻² energy density with a 4.0 mW cm⁻² power density. It also maintained 69.7% of the initial values after 2,000 cycles in the cycling stability tests. Li et al. combined NiO nanoparticles on a hexagonal Ni-MOF through a simple self-template method [192]. The maximum specific capacitance was calculated to be 1192.7 F g⁻¹ at 0.5 A g⁻¹ with 93.32% original values retained after 5,000 cycles. In the assembled Ni-MOF@NiO//AC device, a maximum energy density of 62.2 W h kg⁻¹ was achieved. Xu et al. successfully fabricated ultrathin Cu-MOF@δ-MnO₂ nanosheets in neutral environments via a simple reaction (Figure 8d) [188]. Morphology analysis showed a porous and hierarchical structure with micropores and mesopores. Compared with the other samples, this composite delivered 667 F g⁻¹ at 1 A g⁻¹ (Figure 8e). When assembled with AC to fabricate Cu-MOF@δ-MnO₂//AC ASC, the specific capacitance was 340 F g⁻¹ at 1 A g⁻¹. This ASC also exhibited only a 5% drop compared with the initial values after 6,000 cycles (Figure 8f), indicating its excellent cycling performance. These results proved that metal oxides@MOF is a promising material for supercapacitors.

2.3.3 Conductive polymers

Another effective strategy to enhance the conductivity of MOFs is to incorporate pseudocapacitive materials. Conductive polymers (CPs), such as polyaniline (PANI) and polypyrrole (PPy), can serve as a circuit of electron transport in contact with isolated MOF crystals, thus offering extra pseudocapacitance [193,194]. The addition of these polymers helps reduce the number of electroactive sites and increase internal resistance [195]. PANI is an ideal candidate owing to its low cost, simple manufacturing process, superior capacitance performance, excellent conductivity, promising chemical stability, and unique protonic acid doping mechanism [196,197,198,199]. Liu et al. prepared ZIF-67@PANI composites by coating PANI into the ZIF-67 precursor [200]. From the SEM image, the morphology indicated that ZIF-67 grew in situ under the PANI coating (Figure 9a). They also observed that the ZIF-67@PANI-2 sample obtained the best combination of size and dispersion compared with the others. The GCD experiments exhibited the longest discharge times and increased specific capacitance of 224% (Figure 9b), which was 2,497 F g⁻¹ at 1 A g⁻¹ (Figure 9c). Furthermore, the assembled symmetric supercapacitor constructed by ZIF-67@PANI-2 maintained 92.3% capacitance after 9,000 cycles. The supercapacitor achieved 71.1 W h kg⁻¹ energy density with a power density of 504.72 W kg⁻¹.

Figure 9

(a) SEM image of ZIF-67@PANI-2 samples. (b) GCD curves of four materials at the scan rate of 1 A g⁻¹. (c) Specific capacity of ZIF-67@PANI-1 and ZIF-67@PANI-2. Reproduced with permission from [200]; Copyright 2020, Elsevier. (d) The synthesis process of NiCo-MOF@PNTs. (e) TEM image of NiCo-MOF@PNTs. (f) The specific capacitances at different current densities. (g) The cycling stability of NiCo-MOF@PNTs (inset: the last ten cycles). Reproduced with permission from [201]; Copyright 2019, Elsevier.

In addition to PANI, PPy is also an excellent conjugated conductive polymer that acts as a perfect pseudocapacitive additive for electrode materials in supercapacitors owing to its high electrical conductivity, low cost, high biocompatibility, and excellent adhesion to the substrate [202,203,204]. Jiao et al. fabricated a Zn/Ni-MOF/PPy composite by introducing PPy into bimetallic MOFs [205]. In this study, the optimized composite with 0.15 mL PPy added delivered 202 mA h g⁻¹ at 1 A g⁻¹. Furthermore, an ASC assembled with carbon nanotubes-COOH was constructed and exhibited 50.9 W h kg⁻¹ with a power density of 1,338 W kg⁻¹, retaining 78.8% of its initial value at 10 A g⁻¹ over 5,000 cycles. Liu et al. prepared NiCo-MOF nanosheet coating on nanotubes PPy via a facile ultrasonic method (Figure 9d) [201]. The addition of polypyrrole nanotubes effectively improved the conductivity and prevented aggregation of the NiCo-MOF nanosheets. According to the morphology analysis, NiCo-MOF nanosheets were coated around the polypyrrole nanotubes (PNTs) (Figure 9e). Owing to its unique structure, this composite delivered 1,109 F g⁻¹ at 0.5 A g⁻¹ (Figure 9f). Furthermore, the fabricated ASC (NiCo-MOF@PNTs//AC) presented 41.2 W h kg⁻¹ and a power density of 375 W kg⁻¹. This device also retained 79.1% capacitance after 10,000 cycles (Figure 9g). These results indicated that this nanocomposite is a promising choice for supercapacitors.

2.3.4 Other compounds

Yue et al. explore nonmetallic element (N, S, P, Se, etc.) modified MOFs in supercapacitors for the first time [206]. There is an increasing amount of research on nonmetallic materials applied to supercapacitors. Transition metal phosphides and other chalcogenides have received significant research interest and are utilized in pseudocapacitor electrodes owing to their superior physical and chemical properties [207,208,209]. Yu et al. prepared hierarchical Ni/P/N/C composites using MOF as a template [210]. In the electrochemical tests, Ni/P/N/C-500 (annealing at 500°C) could reach 2887.87 F g⁻¹ at 1 A g⁻¹ with a 6 M KOH solution, and the cycling test presented 90% initial values retained after 1,000 cycles at 10 A g⁻¹. Li et al. fabricated the N, S co-doped bimetal nanocomposite using a Ni/Co-MOF as the precursor [211]. In this study, the specific capacitance reached 1,529 F g^-1 at 1 A g⁻¹, while maintaining 89.29% of the original rate after 10,000 cycles. In an ASC assembled with AC, 41.04 W h kg⁻¹ with a power density of 750 W kg⁻¹ was achieved.

He et al. prepared a functional heterostructure NiCoP-MOF nanosheet-stacked lamellar bricks via hydrothermal-localized phosphorization method [212]. The composites combine the intrinsic merits of MOFs and phosphides, enabling higher activity and fast reaction kinetics. Therefore, a high Faradaic capacitance and rate capability could be enhanced. In three-electrode setup with 2 M KOH, this composite exhibited an outstanding capacitance of 728 C g⁻¹ at 1 A g⁻¹. Furthermore, the device (AC//NiCoP-MOF) achieved a maximum energy density of 50.3 W kg⁻¹ with a power density of 1011.2 W kg⁻¹. It also maintained almost no capacitance loss during 10,000 cycles, proving its excellent cycling stability. Apart from nonmetallic element-related composites, there are other special hybrids that deliver excellent electrochemical performance. Qu et al. prepared Ni-MOF/MXene hybrid nanosheets via the ultrasonic method [213]. In this study, MXene nanosheets are dispersed on the surface of Ni-MOF uniformly, which is a benefit for enhancing its electronic conductivity and preventing the aggregation of Ni-MOF nanosheets. This hybrid showed 867.3 F g⁻¹ specific capacitance at 1 A g⁻¹ and 87.1% capacitance retention after 5,000 cycles at 5 A g⁻¹. Ramachandran et al. synthesized a Co-MOF/MXene hybrid directly on Ni foam [214]. The morphology analysis indicated that the unique interlayer voids space of MXene and Co-MOF on Ni foam could shorten the diffusion path and promote ion intercalation. In 6 M KOH, this composite could achieve an ultrahigh specific capacitance value of 3,741 F g⁻¹ at 3 mA cm⁻², and 92.1% capacitance remained after 3,000 cycles. Liu et al. reported a facile route for preparing NiCo-MOF/acetylene black composites [215]. In the electrochemical measurements, NiCo-MOF/AB-5 achieved 916.1 F g⁻¹ at 1 A g⁻¹. In an assembled ASC, it achieved 33.84 W h kg⁻¹ with a power density of 750 W kg⁻¹. It also maintained 85.25% of the original values after 5,000 cycles, proving its outstanding cycling stability.

Remarkable works on MOF composites as electrode materials for supercapacitors are listed in Table 3.

3.3.1 Graphene

Owing to the excellent performance of graphene and its derivatives, including promising conductivity, high surface area, and chemical stability, they are perfect candidates for combination with PB/PBAs [238]. To enhance their electrochemical activity and conductivity, graphene has been frequently used for fabricating nanocomposites.

A ternary nanocomposite PPy-PB-GO (GO: graphene oxides) was synthesized via spontaneous polymerization among pyrrole, PB, and GO [239]. The scheme could be divided into two stages: (i) absorbing pyrrole onto the GO surface and (ii) using pyrrole to reduce Fe³⁺ to Fe²⁺ and form insoluble PB nanocubes (Figure 12a). The SEM image showed that PB was anchored on the GO matrix (Figure 12b), and the PB nanocubes were protected by PPy. In the electrochemical analysis, the PPy-PB-GO electrodes exhibited 525.4 F g⁻¹, which was considerably better than those of most PBA-based materials and porous carbon materials. The cycling stability tests showed that the PPy-PB-GO electrode could retain 96% of its initial capacitance after 2,000 cycles (Figure 12c), indicating its perfect cycling stability and a high degree of reversibility. Goda et al. synthesized Cu-PBA/GO composites via depositing Cu-PBA nanocubes on GO surfaces [238]. This nanocomposite delivered 611.6 F g⁻¹ at 0.5 A g⁻¹ and maintained 86% primitive values during 2,000 cycles at 4 A g⁻¹. In addition, the assembled supercapacitors could achieve 54.37 W h kg⁻¹ at a power density of 200 W kg⁻¹, proving its outstanding electrochemical performance.

Figure 13

(a) Schematic illustration of the fabrication of CSN-PB/MnO₂ composite. (b) SEM image of CSN-PB/MnO₂ composite. (c) CV curves of CSN-PB/MnO₂ composite at various scan rates. (d) GCD curves of CSN PB/MnO₂ composite electrode at various current densities from 1 to 30 A g⁻¹. (e) The specific capacitance of CSN-PB/MnO₂ composite electrode. (f) Cycling performance of the Fe-PBA/MnO₂//AC device at a current density of 20 A g⁻¹ for 20,000 cycles. Reproduced with permission from [244]; Copyright 2019, Springer. (g) Schematic illustration of the synthesis of PB@MnO₂. (h) SEM image of PB@MnO₂ hybrid. (i) GCD curves of PB@MnO₂ at several current densities. (j) The specific capacitance of PB and PB@MnO₂. (k) Cycling performance of the device at the current density of 1 A g⁻¹. Reproduced with permission from [46]; Copyright 2017, The Royal Society of Chemistry.

Wang et al. fabricated interlocked Graphene@PB nanosheets via a one-step facile hydrothermal technique [240]. They used [Fe(CN)₆]⁴⁻ to release Fe²⁺ and oxidize partial Fe²⁺ to Fe³⁺ in acidic media. Subsequently, the graphene nanosheets could constantly absorb Fe³⁺ ions to react with [Fe(CN)₆]⁴⁻, thus forming PB nanotubes on the graphene nanosheets (Figure 12d). A typical SEM image showed that the PB nanocubes covered both sides of the graphene nanosheet surfaces, forming thicker graphene@PB-5 nanocomposites (Figure 12e). In the electrochemical test, this nanocomposite exhibited 445 F g⁻¹ at 0.5 A g⁻¹, displaying better performance than that of other samples at various current densities (Figure 12f). The assembled supercapacitors delivered 74.8 W h kg ⁻¹ and 5.5 kW kg⁻¹ energy density/power density, respectively (Figure 12g). The nanocomposite also maintained a 97.2% initial value over 5,000 cycles (Figure 12h). This can be ascribed to the synergistic effect of the hybrid nanostructure that prevents the aggregation of PB nanoparticles and restacking of graphene nanosheets.

Sookhakian et al. prepared PB-nitrogen-doped graphene nanocomposites using the Fe(NO₃)₃ solution and K₄Fe(CN)₆ via a simple process [222]. In a 1 M H₂SO₄ solution, this composite-based electrode delivered 660 C g⁻¹ at 1 A g⁻¹. In addition, the stability tests showed that the nanocomposite retained 84.7% of its primitive values during 1,500 cycles, which is also ascribed to the synergetic effects of N-graphene and PB nanoparticles that prevent mechanical deformation. Zhang et al. fabricated PB/rGO hydrogels via a facile one-pot hydrothermal route (denoted as PB-Co/rGOH) [241]. In their study, the CV tests of this hybrid and other compared samples were analyzed, showing that the capacitance of PB-Co/rGOH was mainly attributable to the EDLC and pseudo-capacitance region. Moreover, this composite-made electrode exhibited the longest discharge time and showed 225.2 F g⁻¹ at 1 A g⁻¹ with 82.4% original values maintained over 10,000 cycles. Azhar et al. fabricated iron oxide/carbon composites with nanoporous structures [236]. The morphology clearly showed that the iron oxide nanoparticles are firmly aggregated to each other on the GO nanosheet surface. The specific capacitance value of the iron oxide/carbon composite was 415.0 F g⁻¹ at 2 A g⁻¹, and the capacity retention reached up to 120% over 8,000 cycles. This could be ascribed to the activation of the hybrid that improve the surface wetting of the electrode and led to the fast diffusion of ions in the electrolyte. Xu et al. designed cobalt–nickel hexacyanoferrate hybrids with CNT (K_2.25Ni_0.55Co_0.37Fe(CN)₆/CNT) via a one-step co-precipitation synthesis route [242]. In electrochemical tests, cobalt–nickel hexacyanoferrate hybrids displayed 600 F g⁻¹ at 0.2 A g⁻¹ and maintained 94% of its original value during 2,000 cycles. When assembling this cobalt–nickel hexacyanoferrate with CNT to test its applications, the hybrid electrode achieved 30 W h kg⁻¹ energy density and 60 W kg⁻¹ power density at 0.2 A g⁻¹.

3.3.2 Metal compounds

Transition metal compounds (oxides, hydroxides, etc.) have also been investigated as perfect functional materials for supercapacitors. The hybridization between PB/PBAs and 2D metal oxides leads to the formation of bidirectional coordinative interactions between them, thus enhancing their electrochemical performance [243]. Previous research has shown that PB/PBAs fabricated with MnO₂ can enhance electrochemical performance [237]. MnO₂-based materials are one of the most promising reversible, environmentally benign, and low-cost electrode materials for supercapacitors. They also deliver higher capacitance performances compared with those carbonaceous counterparts [46]. Zhang et al. combined PB with a unique coordination superamolecular network (CSN) with MnO₂ to fabricate CSN-PB/MnO₂ hybrids (Figure 13a) [244]. The SEM image showed that the typical PB was uniformly inserted into MnO₂, and some adhered to the surface of MnO₂, which is favorable for promising electrochemical performance (Figure 13b). In electrochemical tests, this composite electrode exhibited a characteristic pseudocapacitive feature at different scan rates from 10 to 50 mV s⁻¹ (Figure 13c). In further studies, the GCD tests of this composite displayed highly symmetrical properties and relatively linear slopes (Figure 13d). Furthermore, in 1 M Na₂SO₄ solution, this composite electrode could deliver 315.3 F g⁻¹ at 1 A g⁻¹, which was higher than that of pristine MnO₂ and the CSN-PB electrode. In particular, this composite retained 32.4% of the highest capacitance values at 30 A g⁻¹, indicating good rate capability (Figure 13e). Furthermore, the fabricated ASC, CSN-PB/MnO₂//AC retained 85.5% of the original values over 20,000 cycles at 20 A g⁻¹ (Figure 13f). Moreover, this device achieved a maximum energy density of 46.13 W h kg⁻¹ and a power density of 1.20 kW kg⁻¹. Its power density could be as high 18.00 kW kg⁻¹, while retaining an energy density of 19.95 W h kg⁻¹. These performances demonstrated their potential applications in supercapacitors.

Figure 14

(a): 3D diagrams of interlayer C—H···N H-bonding in TpOMe-DAQ. (b) SEM image of TpOMe-DAQ thin sheet. (c) Cycling stability of TpOMe-DAQ at 10 mA cm⁻². Reproduced with permission [52]; Copyright 2018, American Chemical Society. (d) Synthetic scheme of Ni-COF. (e) SEM image of Ni-COF. (f) TEM image of Ni-COF. (g) GCD curves at different current densities (1–10 A g⁻¹). (h) The specific capacitance of Ni-COF and Ni₀-COF. (i) Cycling stability at a current density of 1 A g⁻¹. Reproduced with permission from [53]; Copyright 2019, The Royal Society of Chemistry.

Das et al. [46] fabricated a PB@MnO₂ hybrid electrode material, and this is the first report of a PB composite with metal oxides (Figure 13g). In the morphological analysis, the FE-SEM image showed that MnO₂ is preferentially deposited on the surface of PB nanocubes, offering a porous surface morphology (Figure 13h). To evaluate its electrochemical properties, GCD experiments were conducted at different current densities (Figure 13i). The specific capacitance of PB@MnO₂ was 608 F g⁻¹ at 1 A g⁻¹. Moreover, approximately 68% of the initial values of the PB@MnO₂ hybrid were preserved even at 10 A g⁻¹, indicating promising rate capability compared to pure PB (Figure 13j). Moreover, they assembled a PB@MnO₂ hybrid and polyaniline/graphene nanoplatelet (PG) as ASC. The ASC exhibited great capacitive retention after 4,000 cycles (Figure 13k). Furthermore, it exhibited a 16.5 W h kg⁻¹ energy density along with a 550 W kg⁻¹ power density while retaining 11.3 W h kg⁻¹ energy density under 5,497 W kg⁻¹ power density. These remarkable behaviors demonstrated that the PB@MnO₂ composite is a potential candidate for electrode materials in supercapacitors. Sun et al. designed and synthesized ZnCo-PBA@α-Co(OH)₂ composites by loading α-Co(OH)₂ nanosheets onto Zn₃[Co(CN)₆]₂ nanospheres [245]. In the electrochemical analysis, this composite exhibited 423.92 F g⁻¹ at 1 A g⁻¹. In addition, they constructed a ZnCo-PBA@α-Co(OH)₂//AC ASC, which displayed a 73.53% original value maintained after 5,000 cycles at 5 A g⁻¹, demonstrating its promising cycling stability. The energy density of the ASC was calculated to be 49.13 W h kg⁻¹ with 1734.0 W kg⁻¹ power density, indicating that the device is an excellent electrochemical material.

Remarkable works on PB/PBAs as electrode materials for supercapacitors are listed in Table 4.