Hydrothermal regulation of MnO₂ on a wood-based RGO composite for achieving wide voltage windows and high energy density supercapacitors

Summary

The increasing need for improved energy storage devices renders it particularly important that inexpensive electrodes with high capacitance, excellent cycling stability, and environment-friendly characteristics are developed. In this study, a wood-derived carbon@reduced graphene (WRG) conductive precursor with an average conductivity of 15.38 S/m was firstly synthesized. The binder-free WRG-MnO₂ electrode was successfully constructed by growing MnO₂ onto a WRG under hydrothermal conditions. The asymmetric supercapacitor assembled with the WRG-₂₀MnO₂ cathode exhibited excellent electrochemical capacitive behavior with a voltage window of 0–2 V, maximum energy density of 52.3 Wh kg⁻¹, and maximum power density of 1642.7 W kg⁻¹, which is mainly due to the distinctive icicle-shaped structure of the MnO₂. Thus, a facile strategy for developing high-performance hierarchical porous carbon electrodes that can be used in supercapacitors was developed herein, which may provide new opportunities to improve the high added value of poplar wood.

Graphical abstract

Highlights

• •The concept of biomass electrodes based on wood, RGO, and MnO₂ is firstly proposed
• •The icicle-shaped MnO₂ of WRG-MnO₂ accelerates electrons migration and ions diffusion
• •The packaged WRG-MnO₂/AWC shows excellent overall device performance

Natural sciences; Applied sciences; Materials science

Introduction

The critical shortage of fossil resources and increasingly severe environmental pollution renders the development of sustainable and low-cost energy storage devices increasingly urgent.¹ Supercapacitors (SCs), with advantages such as fast charging and discharging, a long lifespan, and high power density, are therefore particularly attractive green energy storage devices for use in fields such as solar-power generation, wind-power generation, and hybrid vehicles.^2,3 However, the low energy density of SCs limits their application in automobiles and consumer electronics.^4,5 Increasing the voltage window is one effective means of improving the energy density of SCs. The storage energy capability of electrochemical energy storage devices, such as SCs and lithium-ion batteries, also strongly relies on the properties of the electrode material, including its micromorphology, conductivity, and specific surface area.).^6,7,8,9 The good electrical conductivity and high chemical stability of carbon-based electrode materials, such as graphene or porous carbon, render them promising candidates for SCs,¹⁰ and recent advances have led to the development of numerous new electrode materials, such as carbon materials,^11,12 metal oxides,^13,14,15 and conductive polymers,^16,17,18 that can further improve the performance of SCs. These electrode materials are typically in powder form and thus binders and conductive additives are required to obtain precursor ink, which is then coated onto current collectors as substrates, increasing the weight of the obtained electrode without increasing the energy density. However, the coating process can result in inferior electrode structures that limit fast electronic and ionic conduction.¹⁹ In this context, the primary challenge of electrode design is achieving a wide voltage window while maintaining low tortuosity, deformability, and high ionic and electronic conductivity.²⁰

Wood has a unique three-dimensional (3D) hierarchical porous structure that comprises large numbers of vertical vessels (30–130 μm) running along the growth direction of the tree, through which ions, water, and other nutrients are transported around the tree.^21,22 This porous structure mainly comprises tracheids, vessels, and cellulose fibers comprising different arrangements of lignin and hemicellulose and is particularly conducive to the rapid transport of ions and electrons.²³ The natural mechanical structure that is formed by cellulose, hemicellulose, and lignin is similar to reinforced concrete, rendering it a natural precursor for independent carbon electrodes;²⁴ the 3D hierarchical network increases the interaction between electrolyte and electrode surface while also offering a large surface area for the deposition of active materials.^{25,26,27,28,29} Therefore, several researchers have investigated the use of conductive carbonized wood as an electrode material for SCs. However, as a type of carbonaceous material, the limited energy storage capacity that is associated with carbonized wood hinders its energy density,³⁰ and the oxygen-sensitive functional groups on the surface of carbonized wood are highly active and reduce the energy storage performance and cyclic stability of a produced SC.^20,31 Thus, the power and energy density of wooden energy storage devices are often improved by introducing other components, such as metal oxides, polymers, and heteroatoms.^20,32,33

MnO₂ is a typical electrode pseudocapacitor that has characteristics such as high specific capacitance (theoretically reaching up to 1370 F g⁻¹ based on a one-electron redox reaction per manganese atom),³⁴ low cost, and satisfactory environmental compatibility. However, the poor electronic conductivity of MnO₂ (∼10⁻⁸ S cm⁻¹) hinders its practical application in energy devices.³⁵ To address the intrinsic limitations and maximize the utilization of MnO₂ as a pseudocapacitor, one common strategy is to integrate low-dimensional oxide materials such as MnO₂ with highly conductive substrates, such as carbon materials.^36,37,38 Wan et al. prepared a free-standing and binder-free SC electrode by compounding wood biochar with MnO₂,³² with results showing a moderate specific capacitance of 101 F g⁻¹ at a current density of 0.05 A g⁻¹, excellent Coulombic efficiency of 98%–100%, and favorable cyclic stability with a capacitance retention of 85.0% after 10,000 cycles. Further, carbon materials, such as carbon nanotubes (CNTs),^39,40 graphene,^41,42,43 and porous carbon,^44,45,46 exhibit better cycling stability, higher rate capabilities, and wider working voltages than pseudocapacitive materials. Graphene sheets are promising carbon materials for use as SCs owing to their high specific surface area, superior conductivity, and stable lattice structure.^47,48,49,50 However, the stacking effect of graphene hinders the transport of ions and the infiltration of electrolytes into the electrode material, reducing its actual specific capacity. Wu et al. fabricated a wood-derived, porous ACW@RGO electrode by infiltrating graphene oxide (GO) into the directional pores of wood,⁵¹ with the resulting electrode exhibiting an outstanding areal specific capacitance of 26.6 F cm⁻² at a current density of 1 mA cm⁻².

In line with previous research, this study aimed at successfully synthesizing a wood-derived carbon@RGO-MnO₂ (WRG-MnO₂) electrode using a simple hydrothermal method. RGO was uniformly distributed within the 3D pores structure of carbonized wood as continuous layers, forming interconnected conductive network structures upon which MnO₂ could grow evenly. The unique icicle-shaped MnO₂ crystal structure was found to greatly accelerate the transport of ions. The electrochemical performance of the newly constructed WRG-MnO₂ electrode was characterized using capacitance as a measurement indicator. An asymmetric supercapacitor (ASC) was constructed with activated wood-derived carbon (AWC) as the anode material, and its overall performance was further analyzed under the two-electrode system.

Results

The overall synthesis process of the WRG-MnO₂ electrode is shown in Scheme 1. First, wood was processed into thin pieces and GO solution was infiltrated into the pores of the wood using the vacuum impregnation method. Conductive graphene was then grown onto the wood via a combination of ascorbic acid reduction and thermal reduction. The resulting self-standing WRG-MnO₂ electrode formed without the use of any adhesive and conductive agent was successfully prepared using the simple hydrothermal method. The electrode exhibited the following unique advantages: (1) The WRG conductive precursor retained the 3D multilevel straight-through channel network structure of the wood, which is an ideal collector material for pooling currents. (2) The excellent conductivity of RGO was conducive to improving the rate performance of the electrode. (3) The unique icicle-shaped MnO₂ crystals that were synthesized through hydrothermal regulation greatly improved the specific capacitance of the electrode.

Scheme 1.

Schematic illustration showing the preparation process of the WRG-MnO₂ electrode

Analysis of WRG conductivity

The conductivities of the WRG precursors were studied to prove that the WRG can be used as a self-standing conductive fluid collector. The carbon content of GO increased to ∼90% following high-temperature thermal deoxidation; however, the conductivity remained low because of the residual defects on the graphene sheet. Chemical reduction in the form of treating wood@GO with ascorbic acid (3 g/L) over 3 h was therefore performed to improve the electrical conductivity of the obtained WRG. As shown in Figure 1A, the conductivity of the WRG precursors increased from 0 S/m for wood@GO to 0.37 S/m for WRG treated with ascorbic acid at a concentration of 0–5 g/L. This result indicates that GO was reduced and that the defects on the graphene sheet were removed to a certain extent. On this basis, defect removal was further performed through high-temperature heat treatment. The WRG precursor demonstrated the best conductivity (15.38 S/m) using 4 g/L of the GO solution during carbonization (2 h, 1,000°C), which is ∼40% higher than that of the WRG precursor obtained via ascorbic acid reduction (Figure 1B). This reveals that heat treatment further improves the final improvement effect of the graphene structure.

Figure 1.

Conductivity of the WRG precursor

(A) Conductivity changes under VC reduction conditions.

(B) Conductivity changes under carbonization conditions.

Characterization of WRG-MnO₂ as the cathode material

The morphologies and structures of different materials were characterized via SEM and TEM. Abundant 3D hierarchical pore structures (macropores, mesopores, and micropores) were observed within the wood, as seen in Figures 2A and 2B). These long, straight channels provided a natural channel structure for the uniform distribution of layered graphene. Interestingly, the WC still retained the structural advantage of graded wood porosity, facilitating the rapid transport of electrons and electrolyte ions. Uniform layers of graphene in the WC channels (Figure 2C) were in stark contrast with the icicle-shaped MnO₂ crystals that were observed in the RGO along the inner wall of the wood conduit as a result of hydrothermal regulation (Figure 2D). Importantly, as the concentration of KMnO₄ increased, the morphology also underwent considerable changes, with the initially-generated MnO₂ exhibiting a lamellar structure (Figures S1A–S1C); however, as the amount of MnO₂ increased, the structure of the MnO₂ changed to form small spheres with uneven surfaces (Figures S1D–S1F), followed by convex icicle shapes, which resulted from the self-assembly behavior of the clusters (Figures 2E–2G), and ultimately large spherical shapes with smooth surfaces (Figures S1G–S1I)). The icicle-shaped MnO₂ crystal structure can greatly shorten the electron transfer path, which is conducive to rapid electron transfer. An interesting phenomenon was also observed in Figure 2H, in which RGO film formed spontaneously within the wood conduit via self-assembly. Moreover, nano-MnO₂ crystals grew uniformly on the thin RGO films, indicating that the self-assembly behavior of the graphene increased the growth area of the MnO₂, which was particularly conducive to improving the specific capacitance of the electrode. This is mainly due to the oxidation–reduction reaction that occurs between carbon and KMnO₄ under hydrothermal conditions (Equation 1), which allows MnO₂ to grow on the graphene without shedding. TEM observation further confirmed that nano-MnO₂ crystals had grown in situ on the wood carbon and graphene films via the hydrothermal reaction (Figures 2I and 2J)) and that nano-MnO₂ crystals exhibited a long needle shape, which was consistent with the SEM results and further confirmed the in situ growth of MnO₂ on the graphene. The crystal structure of the WRG-₂₀MnO₂ electrode was then studied through HR-TEM. Figures 2K and 2L) show the parallel interatomic planar spacing of 7.1, 2.5, and 1.48 Å, which correspond to the typical (002), (006), and (119) crystallographic planes of the birnessite-type MnO₂ phase, respectively. The three concentric circles in the electron diffraction (SAED) results in Figure 2L indicate the polycrystalline structure. The above results indicate the successful loading of MnO₂ onto the WRG precursor, while the EDX mapping images of the WRG-₂₀MnO₂ electrode also prove that C, O, and Mn are uniformly present within the structure of the WRG-₂₀MnO₂ electrode, with both C and O covered by large amounts of Mn (Figures 2M–2P).

Figure 2.

Microstructural characterizations of materials

(A and B) SEM images of WC: (A) Transverse section and (B) longitudinal section.

(C) Cell wall of the WRG precursor.

(D) Cell wall of the WRG-₂₀MnO₂ electrode.

(E–H) SEM images of the WRG-₂₀MnO₂ electrode at different magnifications.

(I and J) TEM images of the WRG-₂₀MnO₂ electrode.

(K and L) HR-TEM images of the WRG-₂₀MnO₂ electrode, with SAED pattern inset.

(M–P) Element mapping images of the WRG-₂₀MnO₂ electrode: (M) Total spectrum, (N) C, (O) O, (P) Mn.

The crystalline structure of the samples was further studied via XRD. As shown in Figure 3A, the XRD patterns of the WC and WRG showed two wide peaks at 26° and 44°, which corresponded to the (002) and (100) diffraction planes of the graphite layer structure, respectively. The WRG-₂₀MnO₂ electrode exhibited new characteristic peaks at 12.3°, 36.8°, and 65.7°, corresponding to the (002), (006), and (119) diffraction peaks of the birnessite-type MnO₂ (JCPDS no. 18–0802) crystal phase, respectively. In general, birnessite-type MnO₂ nanostructures can be coated in situ onto the outer surface of various carbon materials by redox deposition, creating an intimate connection with the conducting carbons.^52,53 Raman spectroscopy was used to evaluate the defect degree of the electrode, as seen in Figure 3B, with WC, WRG, and WRG-₂₀MnO₂ exhibiting characteristic peaks indicating atomic carbon crystals; the D peak at 1340 cm⁻¹ indicated a defect in the C atom lattice, while the G peak at 1580 cm⁻¹ represented an in-plane stretching vibration of the C atom sp² hybrid. The intensity ratio of the D to G peak (I_D/I_G) represented the degree of disorder in the carbon materials.⁵⁴ Introducing the RGO led to a decrease in the ratio of I_D/I_G from 1.2 to 1.0, indicating a decline in the disordered structure of the material. The Raman bands obtained for MnO₂ were attributed to the Mn–O stretching vibration in the [MnO₆] octahedra (501 cm⁻¹), Mn–O stretching in the basal plane of the [MnO₆] sheets (575 cm⁻¹), and symmetric Mn–O stretching of the [MnO₆] group (645 cm⁻¹).⁵⁵ A strong characteristic peak indicating MnO₂ at 645 cm⁻¹ in the WRG-₂₀MnO₂ spectrum indicated that the MnO₂ was successfully synthesized in the WRG-₂₀MnO₂ electrode. XPS was then used to further verify the chemical state of the Mn in the WRG-₂₀MnO₂ electrode. The survey spectrum in Figure 3C indicated that the WRG-₂₀MnO₂ electrode contained C, O, and Mn. Figure 3D shows the high–resolution XPS spectra of Mn 2p, the spectra was distributed into two different peaks (Mn³⁺ and Mn⁴⁺) at the binding energy of 653 and 653.3eV for Mn 2p_1/2 and 641.3, 641.8eV for Mn 2p_3/2, respectively. In addition, to evaluate the contribution of Mn³⁺ and Mn⁴⁺, the ratio of peak area was used to determine the relative proportion of Mn³⁺ and Mn⁴⁺. The results show that the molar ratio of Mn³⁺ and Mn⁴⁺ was 2:5, and the spin–orbit coupling separation energy of Mn 2p_1/2 (653.7eV) and Mn 2p_3/2 (642 eV) was 11.7 eV, proving that the oxidation state of Mn in the WRG–₂₀MnO₂ electrode material is Mn⁴⁺ (MnO₂).⁵⁶

Figure 3.

Chemical analysis of materials

(A) XRD patterns of WC, WRG, and WRG-₂₀MnO₂.

(B) Raman patterns of WC, WRG, and WRG-₂₀MnO₂.

(C and D) XPS spectra of the WRG-₂₀MnO₂ electrode: (C) Survey XPS and (D) high-resolution Mn 2p spectra.

(E) N₂ adsorption–desorption isotherms of WRG-MnO₂.

(F) Pore diameter distribution of WRG-₂₀MnO₂.

Furthermore, the N₂ adsorption–desorption curves of the WRG-MnO₂ electrode further reflected the changes in the specific surface area and pore size distribution after MnO₂ loading. As shown in Figure 3E, according to the IUPAC classification,⁵⁷ the nitrogen adsorption/desorption isotherm of the WRG-MnO₂ electrode was a typical IV type, indicating a mesoporous structure. It is worth noting that the WRG-₂₀MnO₂ electrode had the largest specific surface area of 98.93 m² g⁻¹ (Table S1). The WRG-₂₀MnO₂ electrode has a large specific surface area as well as is a mesoporous material that allows the easy diffusion of electrolyte ions, which pass through the electrode and electrolyte interface.⁵⁸ In addition, the pore size distribution of the WRG-MnO₂ electrode material was estimated using BJH analysis, with the results in Figure 3F also demonstrating mesoporous (2–50 nm) characteristics and an average pore diameter of 5.9 nm for the WRG-MnO₂ electrode (Table S1). The high specific surface area, active reaction sites, and higher transport efficiency of the reactants in the mesoporous structure render the electrode more conducive to energy conversion and storage.⁵⁹

To further enhance the charge storage capacity of the wood-based thick electrodes, the entire electrode thickness was designed to be 570 μm (Figure 4A). The amount of MnO₂ loaded was controlled by the concentration of KMnO₄ under the specific reaction equation (Equation 1)). As seen in Figure 4B, the amount of MnO₂ loaded increased with the addition of KMnO₄, with 7.14, 13.19, 27.38, and 40.26 mg cm⁻² observed for the WRG-₅MnO₂, WRG-₁₀MnO₂, WRG-₂₀MnO₂, and WRG-₃₀MnO₂ electrodes, respectively. CV was conducted on the electrodes in the three-electrode system for a more effective evaluation of the electrochemical performance of the WRG-MnO₂ electrodes. Figure 4C shows the CV curve at a scan rate of 1 mV s⁻¹ and potential range of 0–0.8 V. It is apparent from the figure that the area enclosed by the CV curves gradually rises as the MnO₂ loading increases. This result directly proves that the MnO₂ content affects the charge storage capacity of the electrode. Of the electrodes tested, the closed area of the WRG-₂₀MnO₂ electrode was obviously improved as compared to the other electrodes.

$${4\text{MnO}}_4^{-}+3\mathrm{C}+{\mathrm{H}}_2\mathrm{O}\to {4\text{MnO}}_2+{\text{CO}}_3^{2-}+{2\text{HCO}}_3^{-}\text{.}$$ (Equation 1)

Figure 4.

Electrochemical characterizations of WRG-MnO₂ electrodes

(A) Thickness measurement of the WRG-MnO₂ electrode.

(B) Different KMnO₄ concentrations correspond to the areal mass loading of MnO₂ in the WRG-MnO₂ electrode.

(C) CV curves of the WRG-MnO₂ electrode at 1 mV s⁻¹.

(D–F) Electrochemical performance of the WRG-₂₀MnO₂ electrode: (D) CV curves obtained at different scan rates. (E) GCD curves obtained at different current densities. (F) Rate performance.

(G) Nyquist diagram of the WRG-₅MnO₂, WRG-₁₀MnO₂, WRG-₂₀MnO₂, and WRG-₃₀MnO₂ electrodes (illustrated as a partially enlarged Nyquist diagram).

(H) Cycling performance of the WRG-₂₀MnO₂.

The electrochemical performance of the prepared electrode was further evaluated in the three-electrode system. As seen in Figure 4D, the WRG-₂₀MnO₂ electrode demonstrated near-mirror symmetry characteristics and there was no obvious redox peak, reflecting typical double-layer capacitance characteristics. In addition, the WRG-₂₀MnO₂ electrode had the largest curve closure area compared to the other electrodes (WRG-₅MnO₂, WRG-₁₀MnO₂, and WRG-₃₀MnO₂) (Figures S2A, S2D, and S2G). The GCD curves of all the electrodes were in the shape of an isosceles triangle, which was explained by the ideal double-layer capacitance behavior of the electrode and is consistent with the CV curves of the electrodes (Figures 4E, S2B, S2E, and S2H)). In Figures 4F, S2C, S2F, and S2I), the WRG-₂₀MnO₂ electrode shows excellent rate performance with an areal capacitance (specific capacitance) of 8.22 F cm⁻² (214 F g⁻¹) at 1 mA cm⁻², with the areal capacitance of the WRG-₅MnO₂, WRG-₁₀MnO₂, and WRG-₃₀MnO₂ electrodes at 2.36, 4.02, and 11.9 F cm⁻², respectively. The areal capacitance of the electrode grows continuously as the MnO₂ loading increases; however, it is worth noting that the voltage drop in the GCD discharge curve of the WRG-₃₀MnO₂ electrode is considerably higher than that of any other electrode (at a current density of 1 mA cm⁻², the voltage drop of WRG-₃₀MnO₂ is 0.0501V, WRG-₅MnO₂ is 0.0145V, WRG-₁₀MnO₂ is 0.0291V, and WRG-₂₀MnO₂ is 0.0191V), which may be the result of excessive MnO₂. The agglomeration effect caused by excessive MnO₂ directly hinders the transfer of electrons and ions, increasing the resistance and reducing the conductivity of the electrode. Moreover, the unique icicle-shaped crystal structure of the MnO₂ accelerates the rapid transfer of the electrons, improving the electrochemical performance of the electrode. Compared with similar carbon-based MnO₂ electrodes, the WRG-₂₀MnO₂ electrodes also exhibited outstanding rate performance (Table S2).

Electrochemical impedance was used to evaluate the electronic conductivity and charge transport characteristics of the electrodes in the frequency range of 10⁵–10⁻² Hz at an internal alternating current (AC) disturbance of 5 mV. Figure 4G shows the Nyquist diagrams of the WRG-₅MnO₂, WRG-₁₀MnO₂, WRG-₂₀MnO₂, and WRG-₃₀MnO₂ electrodes, with the corresponding equivalent fitting circuit diagram. The Nyquist diagram is semicircular in the high-frequency region and linear in the low-frequency region. In addition to the double-layer capacitance formation (C_d), the first intercept at high frequency between the semicircle and the transverse axis represents the internal bulk resistance (R_Ω), which includes the ionic resistance of the electrolyte solution and the intrinsic resistance of the active material,⁶⁰ while the diameter of the semicircle reflects the charge transfer resistance (R_ct)⁶¹ and the linear part reflects the Warburg impedance (Z_w), which is caused by the diffusion of electrolytic ions onto the surface electrode. The x-intercept of the high-frequency part of the Nyquist plot shows that the conductivity of the WRG-₅MnO₂, WRG-₁₀MnO₂, WRG-₂₀MnO₂, and WRG-₃₀MnO₂ electrodes is gradually worsening. This occurs because as the loading increases, nano-MnO₂ gradually agglomerates, making its conductivity poor. The excess MnO₂ inhibits the WRG precursor, limiting the conductivity of the WRG precursor, increasing the R_Ω of the electrode, and decreasing the conductivity of the electrode. The WRG-₂₀MnO₂ electrode has the lowest R_ct (4.5 Ω) as compared to WRG-₁₀MnO₂ (8.1 Ω) and WRG-₃₀MnO₂ (19.7 Ω), indicating that WRG-₂₀MnO₂ has considerable advantage for the charge transport capability. In the mid frequency region, the projection length of the Warburg impedance curve on the Z′ axis characterizes the ion penetration process. Obviously, the Warburg length of the WRG-₂₀MnO₂ electrode is shorter than that of other electrodes (WRG-₅MnO₂, WRG-₁₀MnO₂ and WRG-₃₀MnO₂), which indicates that the WRG-₂₀MnO₂ electrode has the shortest ion diffusion path. This is mainly attributed to the icicle shapes nano MnO₂ structure. In addition, the cycling performance of WRG-₂₀MnO₂ was investigated (Figure 4H), with results showing that 85.3% of its initial capacitance was retained after charging and discharging for 5,000 cycles at 10 mA cm⁻², demonstrating its excellent cycling performance. The GCD curve maintained almost the same shape in 20 randomly selected cycles (illustrations for Figure 4H), demonstrating the outstanding reversibility of the WRG-₂₀MnO₂ electrode. The above results demonstrate that the WRG-₂₀MnO₂ electrode has excellent electronic conductivity and charge transfer characteristics, indicating its obvious advantages in the use of SCs with high energy density.

Conductivity and energy storage mechanism of the WRG-MnO₂ cathode material

The conductive and energy storage mechanisms of the WRG-₂₀MnO₂ electrode were also examined in detail to gain a more comprehensive understanding of its working performance. The results indicated that the –OH and –COOH in the GO were chemically bonded to the –OH in the wood to form hydrogen bonds as a result of vacuum impregnation, anchoring GO tightly to the wood. Figure 5 shows the conduction and energy storage mechanism of the WRG-MnO₂ electrode. After filling the wood with RGO, the wood surface and inner microchannel wall were observed to be fully covered by lamellar RGO. Several oxygen-containing groups from wood@GO (–COOH, –OH, C=O, –C–O–C) were removed in the form of H₂O and CO₂ during thermal reduction. The thermal decomposition of the oxygen-containing groups reduced the binding of oxygen atoms to the electrons on the carbon atoms, thereby promoting the rapid migration of electrons onto the carbon plane in the interconnected microchannel structure. The RGO obtained by the reduction endowed the WRG precursor with high conductivity characteristics, and the lamellar graphene provided a large specific surface area for the loading of nano-MnO₂. When WRG and KMnO₄ underwent hydrothermal reaction, the C–Mn bond that formed to produce MnO₂ was connected via bonding, effectively improving the capacitive characteristics of the electrode.

Figure 5.

Diagram showing the conductivity and energy storage mechanism of the WRG-₂₀MnO₂ electrode

Characterization of AWC as the anode material

The importance of the negative electrodes that are used in energy storage devices cannot be ignored. Here, we investigate the effect of activation time on the AWC performance. Analysis of the morphology structure of the AWC electrode demonstrated that increasing numbers of pores appear on the AWC electrode as the CO₂ activation time is prolonged (Figures 6A, 6B, and S3). However, the pores on the AWC-1 electrode are relatively sparse and unevenly distributed. The hole etching in the AWC-3 electrode is relatively serious, as manifested by the increasingly random pore structure. In comparison, the new holes that are etched in the AWC-2 electrode are uniform, with a more regular structure, and exhibit improved microscopic morphology. The CV and GCD curves also demonstrate that AWC-2 has better capacitive performance (Figures 6C and 6D). Figure 6E shows a typical CV curve obtained at a scan rate of 1–20 mV s⁻¹ over a scan window of −1.2–0.8 V. The CV curve shows a nearly rectangular shape even at a high scan rate of 20 mV s⁻¹. Compared with the CV curves for AWC-1 and AWC-3, the CV curve of AWC-2 also exhibits the largest closed pattern area (Figures S4A and S4D), indicating that the AWC-2 electrode has excellent specific capacitance. The GCD curves show a nearly triangular shape for all AWC electrodes at 1–20 mA cm⁻² (Figures 6F, S4B, and S4E). Surprisingly, the AWC-2 electrode shows excellent capacitance performance, with corresponding areal capacitances of 5.8, 4.54, 4.08, 3.58, and 3.13 F cm⁻² at current densities of 1, 2, 3, 5, and 10 mA cm⁻², respectively; moreover, even at an ultrahigh current density of 20 mA cm⁻², the AWC-2 still retains an areal capacitance of 2.76 F cm⁻² (Figure 6G). Compared with the other electrodes (AWC-1 and AWC-3), the AWC-2 electrode also holds the highest areal capacitances (Figures S4C and S4F). The above excellent electrochemical performance obtained for the AWC-2 electrode can be attributed to the following reasons: 1) the AWC-2 electrode has a graded porous structure, which promotes the efficient penetration of electrolytes inside the electrode; 2) the AWC-2 electrode shows low curvature throughout the channel, enabling fast ion transport; 3) the conductive 3D carbon network structure of the AWC-2 electrode ensures rapid electron transport. The Nyquist diagram of the AWC electrode also indicates the excellent electronic conductivity and charge transport characteristics obtained for the AWC-2 electrode (Figure 6H), which is consistent with the above analysis. Figure 6I shows that the GCD curve of AWC-2 remains almost unchanged after 20 charge and discharge cycles.

Figure 6.

Microstructural and electrochemical characterizations of AWC electrodes

(A and B) SEM images of the AWC-2 electrode at different magnifications.

(C) CV curves of the AWC electrode at 1 mV s⁻¹.

(D) GCD curves of the AWC electrode at 1 mA cm⁻².

(E–I) Electrochemical performance of AWC-2: (E) CV curves at different scan rates. (F) GCD curves at different current densities. (G) Rate performance. (H) Nyquist diagram of the AWC-1, AWC-2, and AWC-3 electrodes. (I) Cycling performance of the AWC-2.

Electrochemical performances of the WRG-₂₀MnO₂//AWC-2

One of the characteristics of ASCs is their high energy and power density. The operating voltage window of the ASC can be widened using different combinations of cathodes and anodes at opposite potential windows, rendering the electrode promising for practical application. The WRG-₂₀MnO₂//AWC-2 ASC was constructed using an AWC electrode as the anode, a cellulose membrane as the diaphragm, and a WRG-₂₀MnO₂ electrode as the cathode. The ASC was assembled and measured in a 1-M Na₂SO₄ electrolyte. The ASC model diagram and electron migration path are exhibited in Figure 7A. The natural porous structure of wood provides sufficient exchange space for the migration of electrons and ions. A voltage window of 0–2 V was obtained for the fabricated ASC device. Figure 7B shows the CV curve of ASC at different scan rates (5–100 mV s⁻¹) within the stable voltage range of 0–2 V. Increasing the scan rate led to a gradual increase in the area enclosed by the CV curve, and the rectangular shape is retained at a scan rate of 100 mV s⁻¹, indicating that the device has ideal capacitance behavior. To better understand the energy storage behavior of ASC, the contribution of capacitance to the total stored charge was studied using the CV method. It is known that the generated current mainly comes from capacitance-controlled (i_cap) and diffusion-controlled (i_diff) processes, with the calculation relations shown in Equations 2 and 3. The b values in the equations, which range from 0.5 to 1.0, imply a “transition” area between the diffusion-controlled process and pseudocapacitive behavior.⁶² It is worth noting that a b value of 0.5 indicates diffusion contribution, with smaller values indicating greater diffusion contribution, while a b value of 1.0 indicates capacitance contribution, with larger b values indicating greater capacitance contribution. From Figure 7C, it can be seen that the b values under different voltages are mainly distributed in the range of 0.5–1.0. It demonstrates that the energy storage reaction of ASC is a reaction mechanism under the dual effects of pseudocapacitive behavior and diffusion control. Furthermore, the contribution of the pseudocapacitive behavior appears to be more pronounced than the diffusion control, reflecting the storage behavior of the pseudocapacitive charges. Throughout the entire charging and discharging process, the contribution of pseudocapacitive behavior and diffusion control is quantitatively estimated. Specifically, the capacitive quantitative differentiation of the CV curve is performed according to Equation 4. Figure 7D shows the contribution of capacitance to the total storage charge (rose area) at 50 mV s⁻¹ (see Figure S5 for other scan rates), with results demonstrating considerable capacitance contribution in the CV process of the ASC device. As the scan rate increases from 5 to 100 mV s⁻¹, the total proportion of the stored charge from the capacitive-controlled process gradually grows from 13% to 70% (Figure 7E), implying that capacitance plays a dominant role in the ASC storage electrochemical reaction. This is also one of the reasons for the excellent performance observed.

Figure 7.

Electrochemical behavior of WRG-₂₀MnO₂//AWC-2 devices

(A) Model diagram of the WRG-₂₀MnO₂//AWC-2 ASC and schematic diagram of the electron migration path.

(B) CV curves at various scan rates from 5 to 100 mV s⁻¹ in a potential window of 0–2 V.

(C) Log i vs. log v plots under different voltages.

(D) Capacitive contribution to total current at 50 mV s⁻¹.

(E) Contribution proportion of the capacitive- and diffusion-controlled process at different scan rates.

(F) GCD curves at different current densities from 1 to 100 mA cm⁻² in a potential window of 0–2 V.

(G) Rate performances.

(H) Areal and specific energy/power density.

(I) Cycling performance at 50 mA cm⁻² after 10,000 cycles.

(J) Ragone plots of the ASC device compared with other energy storage technologies.

(K) Comparison of the specific capacitance and voltage window of ASC with other devices.

(L) Comparison of the energy/power density of ASC with previously reported devices.

The calculation relations are as follows:

$$\mathrm{i}={\mathrm{i}}_{\text{cap}}+{\mathrm{i}}_{\text{diff}}={\text{av}}^{\mathrm{b}}\text{,}$$ (Equation 2)

$$\log \mathrm{i}=\text{loga}+\text{blogv,}$$ (Equation 3)

where a and b are the adjustable parameters.

$$\mathrm{I}\left(\mathrm{V}\right)={\mathrm{k}}_1\mathrm{v}+{\mathrm{k}}_2{\mathrm{v}}^{1/2}\text{,}$$ (Equation 4)

where i (V) is the current density response at a fixed potential, k₁ is the capacitive effect factor, k₂ is the diffusion process effect factor, and V is the scan rate. Thus, in Equation 4, k₁v and k₂v^1/2 represent the pseudocapacitance contribution and diffusion-controlled contribution, respectively.

The GCD curves at different current densities are shown in Figure 7F. The approximately symmetrical triangular shape indicates ideal capacitance behavior for the ASC. The areal and specific capacitances of the ASC device were calculated according to the discharge curve (Figure 7G), with results showing areal and specific capacitances of 2.7, 2.4, 2.2, 1.7, and 1.3 F cm⁻² and 96, 86, 79.4, 62, and 48 F g⁻¹ at current densities of 1.0, 5.0, 10.0, 30.0, and 50 mA cm⁻², respectively. It is stimulating that the areal and specific capacitance of ASC can still reach 0.7 F cm⁻² and 25 F g⁻¹, respectively, even at the high current density of 100 mA cm⁻². Energy and power density are also important parameters for the ASC. As seen in Figure 7H, the assembled ASC device displays extremely high energy densities (0.375–1.5 mWh cm⁻² and 13.75–39.4 Wh kg⁻¹) and power densities (9.45–36 mW cm⁻² and 346.5–1320 W kg⁻¹). Finally, the cyclic stability of the ASC devices was also evaluated (Figure 7I), with results showing that 81.4% of the initial capacitance of the WRG-₂₀MnO₂//AWC-2 was retained after 10,000 cycles at a current density of 50 mA cm⁻².

Figure 7J shows a Ragone diagram comparing different electrochemical energy storage technologies. The figure indicates that the WRG-₂₀MnO₂//AWC-2 ASC exhibit better energy and power density as compared to energy storage devices such as electrochemical capacitors, lead batteries (PbO₂/Pb), nickel–hydrogen batteries (Ni/MH), primary lithium batteries, and secondary lithium batteries (Li-primary and Li-ion batteries), effectively filling the gap between traditional capacitors and lithium-ion batteries. To more specifically demonstrate the obvious advantages of the ASC prepared in this study, the performance of the obtained device was compared with other devices reported in the literature. As shown in Figure 7K, the ASC exhibits excellent specific capacitance (96 F g⁻¹) and wide voltage window (0–2.0V), which is superior to many of the previously reported SC devices, including WC@MnO₂-20//WC@MnO₂-20 (87 F g⁻¹, 0–1 V),⁶³ NiCo-LDH@NDC//NDC (79.1 F g⁻¹, 0–1.5 V),⁶⁴ ACW@RGO//ACW@RGO (32.1 F g⁻¹, 0–1 V),⁵¹ MnO₂@CC//TiO₂@CC (72.8 F g⁻¹, 0–2.6 V),⁶⁵ and Ni₃S₂/MnO₂//AC (34.8 F g⁻¹, 0–1.5 V).⁶⁶ In addition, we compared the energy density and power density of the ASC with other devices. As shown in Figure 7L, compared to NiCo-LDH@NDC//NDC (1.48 mWh cm⁻², 22.4 mW cm⁻²),⁶⁴ ACW@RGO//ACW@RGO (1.07 mWh cm⁻², 12.2 mW cm⁻²),⁵¹ MnO₂@CC//TiO₂@CC (0.95 mWh cm⁻², 65.0 mW cm⁻²),⁶⁵ Ni₃S₂/MnO₂//AC (0.04 mWh cm⁻², 2.88 mW cm⁻²),⁶⁶ WGNC//WG (0.75 mWh cm⁻², 22.5 mW cm⁻²),⁶⁷ MnO₂/PANI@AWC//CNT@AWC (0.85 mWh cm⁻², 9.0 mW cm⁻²),⁶⁸ rGO@PANI-NFN/OCC//rGO@PANI-NFN/OCC (0.1179 mWh cm⁻², 8.42 mW cm⁻²),⁶⁹ and MnO₂/PANI/CNT//PANI/CNT(0.5133 mWh cm⁻², 3.98 mW cm⁻²),⁷⁰ the energy and power density (1.5 mWh cm⁻² and 44.8 mW cm⁻², respectively) of the ASC presents a prominent performance advantage. This superiority may be due to the excellent performance of the layered RGO and icicle-shaped MnO₂ in the electrodes during the operation of the WRG-₂₀MnO₂//AWC-2 ASC. Our research results thus suggest broad application prospects for WRG-₂₀MnO₂//AWC-2 in the future.

Discussion

In summary, hydrothermal regulation was used for the successful synthesis of a green WRG-₂₀MnO₂ electrode material without the need for adhesives. WC preserves the porous structure of natural wood, providing a fast pathway for electrons migration and ions diffusion. The high capacitance of the WRG-₂₀MnO₂ electrode (214 F g⁻¹/8.2 F cm⁻² at 1 mA cm⁻²) was achieved through the synergistic action of wood, graphene, and MnO₂. An ASC device (WRG-₂₀MnO₂//AWC-2) was constructed using WRG-₂₀MnO₂ as the cathode and AWC as the anode. Notably, the ASC device exhibits a wide voltage window (0–2 V), providing it with excellent energy density (0.375–1.5 mWh cm⁻² and 13.75–52.3 Wh kg⁻¹) and power density (9.45–44.8 mW cm⁻² and 346.5–1642.7 W kg⁻¹) as well as a long cycling life (with 81.4% of the capacitance retained over 10,000 cycles at 50 mA cm⁻²). These exciting research results are attributed to the porous channels of wood, the excellent conductivity and high specific surface area of graphene, and the unique icicle-shaped nano MnO₂ structure. This study provides a new concept for the development and utilization of high-energy-density SCs while providing support for the application of biomass-based materials in the field of renewable electrochemical energy storage.

Limitations of the study

In this work, the 3D porous structure of wood has made a significant contribution to the excellent electrochemical performance of WRG-MnO2. However, wood is an anisotropic biomass material, and the stability of its performance still poses significant challenges in the research of wood-based energy storage materials. In addition, the structural design and load control of electrode materials will also be the focus of research in the following work, which is crucial for improving electrochemical performance.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
Potassium permanganate	Fuchen (Tianjin) Chemical Reagent Co., Ltd	CAS : 7722-64-7
Sodium sulphate anhydrous	Fuchen (Tianjin) Chemical Reagent Co., Ltd	CAS: 7757.82-6
Ascorbic acid	China National Pharmaceutical Group Chemical Reagent Co., Ltd	CAS: 50-81-7
Anhydrous ethanol	Fuchen (Tianjin) Chemical Reagent Co., Ltd	CAS: 64-17-5
Graphene oxide	individual	N/A

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Ximing Wang (wangximing@imau.edu.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •Data reported in this paper will be shared by the lead contact upon request.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Method details

Materials

Eleven-year-old poplar was selected from a fast-growing plantation as the base material and straight-grained sapwood boards with a 151% initial moisture content were obtained from Hohhot, Inner Mongolia Autonomous Region, China. GO was prepared using the improved Hummers method.⁷¹ All chemicals used in this work were newly purchased and used without further purification and were of analytical grade.

Synthesis of a WRG conductive precursor

Wood samples were immersed in a 4 g/L GO solution, and wood@GO composites were obtained by drying at 53°C for 24 h. The WRG conductive precursor was then constructed using a two-step reduction method. (1) chemical reduction: the WC@GO composites were immersed in an ascorbic acid solution (3 g/L) at 80°C for 3 h. (2) Physical thermal reduction: to further modify the structural integrity of GO, the obtained materials were placed into a tubular furnace in a protective N₂ atmosphere, heated at 5 °C min⁻¹, and stabilized at 260°C for over 2 h. Samples were then carbonized at 1000°C for 2 h before the temperature was decreased to 25 °C at a cooling rate of 10 °C min⁻¹.

Preparation of the WRG-MnO₂ cathode

The WRG-MnO₂ composite electrode was synthesized using a highly scalable hydrothermal method. First, a mixed solution of KMnO₄ and Na₂SO₄ (1:1) was prepared and stirred at room temperature until completely dissolved. The WRG precursor was then immersed in the solution before transferring to a 50-mL Teflon-lined autoclave, where it was reacted at 120°C for 3 h. Samples were then cooled to 25°C, washed multiple times with absolute ethanol and DI, and dried at 53°C for 24 h to obtain the WRG-MnO₂ electrode. To study how MnO₂ affects the electrochemical properties, electrodes with different KMnO₄ concentrations (5, 10, 20, and 30 mM) were prepared and denoted as WRG-₅MnO₂, WRG-₁₀MnO₂, WRG-₂₀MnO₂, and WRG-₃₀MnO₂, respectively.

Fabrication of the AWC anode

Preprepared thin wood slices were placed in a tube furnace in a N₂ atmosphere, precarbonized at 260°C for 2 h, and carbonized at 1000°C for 2 h before cooling to 25°C. The obtained WC was then placed in a tube furnace and physically activated in a CO₂ atmosphere at 800°C before cooling to 25°C to obtain the AWC electrode. Heating and cooling were performed at 5 °C min⁻¹ and 10°C min⁻¹, respectively. The effect of activation time on the electrochemical performance was also studied by subjecting the electrodes to different activation periods (1, 2, and 3 h), and the obtained samples were denoted as AWC-1, AWC-2, and AWC-3, respectively.

Materials characterization

The conductivity of WRG-MnO₂ was tested using a four-probe tester (RTS-8, Britain). The morphologies of the sections were observed via scanning electron microscopy (SEM, S-4800I, Japan) and transmission electron microscopy (TEM, FEI Tecnai G2, America). Material elements were analyzed via energy-dispersive X-ray spectrometry (Bruker QUANTAX 200, Germany). The structure of the material was characterized via X-ray diffraction (XRD, TD-3500, China). The chemical composition of the slices was analyzed via X-ray photoelectron spectroscopy (XPS, AXIS Ultra DLD, Japan). Raman spectra were recorded at 633 nm using a Raman spectrometer (Raman, Jobin Yvon HR-Evolution 2, France). The specific surface area was calculated using the Brunauer–Emmett–Teller (BET) method, and the pore size distribution and pore volume were calculated using the Barrett–Joyner–Halenda (BJH) method.

Electrochemical measurements

The electrochemical properties of the electrodes were evaluated in a 1.0-M Na₂SO₄ electrolyte using a three-electrode test cell with a Chi 660E electrochemical workstation (Shanghai Chenhua Instrument Co., Ltd.). A platinum electrode was used as the counter electrode, an Ag/AgCl electrode was used as the reference electrode, and the synthesized WRG-MnO₂ electrode without polymer binders was used as the working electrode. Cyclic voltammetry (CV) and galvanostatic charge–discharge (GCD) curves were obtained at a voltage window of 0–0.8 V. Electrochemical impedance spectroscopy was performed in the frequency range of 10⁵–10⁻² Hz. The areal (C_s, F cm⁻²) and specific (C_m, F g⁻¹) capacitances were calculated based on the GCD curves using Equations 5 and 6.

$$C_s=\frac{I\Delta t}{s\Delta V}\text{,}$$ (Equation 5)

$$C_m=\frac{I\Delta t}{m\Delta V}\text{,}$$ (Equation 6)

where I (A) is the discharge current, $\Delta$ t (s) is the discharge time, $\Delta$ V (V) is the potential window, s (cm²) is the area of the electrode that was exposed to the electrolyte, and m (g) is the mass of the active materials in the electrode.

Asymmetric supercapacitors (ASCs) were assembled using the WRG-MnO₂ and AWC electrodes, i.e., WRG-₂₀MnO₂//AWC-2. The electrochemical performances of the SCs were evaluated in a two-electrode test system. The energy density (E, Wh kg⁻¹) and power density (P, W kg⁻¹) were calculated based on the GCD discharge curve using Equations 7 and 8.

$$E=\frac{1}{2}C\Delta V^2\text{,}$$ (Equation 7)

$$P=\frac{E}{\Delta t}\text{,}$$ (Equation 8)

where $\Delta$ V(V) is the voltage drop and $\Delta$ t (s) is the discharging time.

Published: February 16, 2024