High-performance aqueous symmetric supercapacitors with a versatile 3D reduced graphene oxide aerogel electrode

Synthetic graphite was acquired from TimCal. Hydrochloric acid (35%, HCl), sulfuric acid (96%, H₂SO₄), hydrogen peroxide (30%, H₂O₂), sodium nitrate (NaNO₃), potassium permanganate (KMnO₄) were of analytical grade obtained from Chempur, Poland. Hydroquinone (HQ), sulfuric acid (1 M H₂SO₄), PVDF (Kynar Flex) were ASC grade delivered by Merck Sigma–Aldrich. The chemical reagents were used as received without further purification.

Graphene oxide (GO) was synthesized through the chemical oxidation of graphite using a modified Hummers method, following previously described procedures [23]. A 75 mL solution of 3.5 mg mL⁻¹ GO dispersion was sealed in a Teflon-lined autoclave without a stirrer and maintained at 180 °C for 24 h. Subsequently, the reactor was naturally cooled to room temperature to produce the self-assembling rGO hydrogel. The product was then transferred to a beaker of Milli-Q water, where the water was replaced multiple times over 20 h to achieve a neutral pH. The resulting hydrogel was freeze-dried for 48 h to remove the redundant water and obtain the rGO aero. A previously studied electrode material was also used to investigate the impact of synthesis conditions on the physical structure of rGO and, consequently, its electrochemical properties. Briefly, for the synthesis of rGO HT, the GO solution underwent a hydrothermal reduction in a stainless-steel autoclave under the same temperature conditions, with suspension agitation. After the reaction, the autoclave was cooled to room temperature. The resulting rGO was washed thrice with Milli-Q water to achieve a neutral pH and then vacuum-dried at 60 °C overnight.

N₂ sorption at 77 K was carried out using an Autosorb IQ gas sorption analyser (Quantachrome). Before the measurements the samples were degassed in vacuum at 120 °C for 20 h. The specific surface area (S_BET), total pore volume, and mesopores volume of both rGO materials were measured. The S_BET area was determined in accordance with the Rouquerol criteria [24]. The total pore volume was determined from the relative pressure point p/p₀ less than 0.955, representing pores with diameter less than 50 nm. Micropores volume was calculated using Dubinin-Radushkevich equation. The mesopore volume of the material was determined from the difference of total pore volume and micropore volume, The pore size distribution (PSD) was determined by the quenched solid density functional theory (QSDFT) from the equilibrium adsorption points. Field emission scanning electron microscopy (FESEM) measurements were conducted using Merlin Zeiss equipment, and high-resolution transmission electron microscopy (HRTEM) observations were performed on FEI Titan G₃ microscope. X-ray diffraction (XRD) analysis was carried out using a Rigaku Ultima IV diffractometer equipped with Cu Kα radiation. Raman spectra were obtained with a Bruker Senterra Raman Spectrometer and thermal stability of the samples was evaluated using a TGA 8000 thermogravimetric analyzer (Perkin Elmer) under a nitrogen atmosphere, with the temperature increasing from 30 to 800 °C. X-ray photoelectron spectroscopy (XPS) analysis was performed using PHI 5000 VersaProbe equipment.

The rGO aero and rGO HT electrodes, composed of 85% active material, 10% binder (polyvinylidene fluoride, Kynar Flex), and 5% carbon black, were pressed and cut into 10 mm diameter discs with a thickness of 100 µm, separated with Whatman® glassy paper and immersed in different aqueous electrolytes (1 M Na₂SO₄, 1 M H₂SO₄, and 1 M H₂SO₄ + 0.2 M hydroquinone (HQ)). The total active mass of electrodes in the investigated symmetric systems was in the range of 12–18 mg, maintaining the equal masses of the positive and negative electrodes. Two gold plates were used as the current collectors, and the electrochemical performance of the device was investigated in a two-electrode Swagelok™-type setup.

Supercapacitor cells performance was evaluated through CV, GCD, and electrochemical impedance spectroscopy (EIS) using a Biologic VSP potentiostat–galvanostat. The potential range for CV and GCD measurements was between 0–2.0 and 0–1.4 V for neutral and redox-doped electrolytes, respectively. For the GCD technique, currents ranging from 0.2 to 10 A g⁻¹ were used to examine the effect of the current load on capacitive behavior. EIS measurements were conducted in the 400 kHz to 10 mHz frequency range at a 5 mV amplitude referring to open circuit potential.

The stored energy density (E, Wh kg⁻¹) and delivered power density (P, kW kg⁻¹) of the symmetric device presented in the Ragone plot were calculated using the equations as follows:where I_m is current density in Ag⁻¹,\(\underset{{U}^{-}}{\overset{{U}^{+}}{\int }}U dt\) is current integral area, and Δt is the discharge time (s) [25, 26].

The gravimetric capacitance of the electrode (C_GCD, F g⁻¹) based on GCD data was calculated using the equation provided, as the charge–discharge profiles exhibit non-linear variation resulting from quasi-reversible faradic reactions [27]:where E is energy density calculated according to the Eq. (1) in Wh kg⁻¹ and U_max is a maximum safe operating voltage in V.

Additional calculation protocols for the deeper analysis of the CV measurements data (coulombic efficiency, energy efficiency, b-value and surface-controlled and diffusion-controlled current contribution) are included in the supplementary information.

The morphology of the rGO aero (Fig. 1a) and rGO HT (Fig. 1b) materials was examined through FESEM observations. The texture obtained in the rGO aero exhibited interconnectivity, with sharp edges and a thin wall thickness. The graphene aerogel sheets displayed variable orientation toward the cross section, resulting in the formation of random openings within the material. Pores exceeding 400 nm in diameter were visible on the surface of the material. This structure is a characteristic of porous graphene materials and aerogels obtained through freeze drying [9, 28]. By contrast, the rGO HT synthesized under conventional hydrothermal conditions displayed a typical aggregated morphology characterized by densely stacked graphene nanosheets [29]. Furthermore, HRTEM images further supports these structural disparities, with rGO aerogel (Fig. 1c) demonstrating well-defined, thin and continuous framework of graphene sheets. In contrast, the conventional rGO HT (Fig. 1d) displays much larger, more opaque regions, indicating significant restacking and aggregation of graphene layers. These observations supports that the aerogel structure enhances the spatial arrangement and achieves interconnected network of graphene sheets.

Bild vergrößern

Figure 2 shows the N₂ sorption isotherms at 77 K and PSD profiles of the graphene materials. Both the rGO aero and rGO HT displayed a Type IV isotherm with a distinctive hysteresis loop (Fig. 2a) according to the IUPAC classification [30]. As follows from Table 1, the rGO aero had a well-developed porous structure with a BET specific surface area of 495 m² g⁻¹, slightly higher than the material obtained through the conventional route [31]. Both rGO HT and the rGO aero exhibited comparable contributions of mesopores in the total pore volume (V_mes/V_t = 0.57 − 0.59), however, the rGO aero displayed a significantly higher total pore volume than rGO HT (0.445 vs. 0.324 cm³ g⁻¹). A comparable proportion of micropores and mesopores appears beneficial for efficient supercapacitor operation [32, 33]. Pore size distribution analysis revealed that the rGO aero exhibited pores across a wide range, whereas rGO HT featured micropores and narrow mesopores with a size up to 6 nm (Fig. 2b).

Bild vergrößern

The calculated average pore diameter, using the 4V_T/S_BET formula based on data from Table 1, suggests larger average pore diameters for both materials (3.60 nm for the rGO aero and 2.86 nm for rGO HT) than the sizes of ions commonly found in aqueous electrolytes [31]. The wider pores of the rGO aero suggest better electrochemical performance by ensuring greater availability of the material’s active sites to the electrolyte ions and faster ionic diffusion. Moreover, the rGO aero has a larger volume of micropores (0.192 vs. 0.133 cm³ g⁻¹), which play an important role in charge accumulation. At the same time, there are well-developed transport channels for ions diffusing from the inside of the electrolyte to the phase interface because of the large number of narrow mesopores. They provide good charge propagation, which allows them to achieve high capacitance values and lower the electrode internal resistance [33].

The X-ray diffractograms (Fig. 3a) of the rGO aero and rGO HT exhibit a prominent diffraction peak at 2θ = 24.3° and 24.7°, respectively, corresponding to the (002) lattice plane of a typical graphitic structures. Additionally, weaker peak at 2θ = 43.1° is observed, which is attributed to the (100) plane [34]. The increase in d₀₀₂ from 0.3604 nm (rGO HT) to 0.3650 nm (rGO aero) results from the minimization of capillary-induced collapse of the graphene structure during freeze-drying, which preserves a more open architecture and increases the spacing between graphene layers [35].

Bild vergrößern

Further structural differences between rGOs synthesized via conventional hydrothermal treatment versus a freeze dried aerogel were investigated by Raman spectroscopy (Fig. 3b). Two prominent bands corresponding to the D-band (~ 1340 cm⁻¹) and G-band (~ 1590 cm⁻¹), characteristic for graphene-based materials are identified. The intensity ratios of these bands I_D/I_G are key metric used to quantify the level of in-plane structural defects and disorder within the graphene lattice [36]. The calculated I_D/I_G ratio for the rGO HT sample is 1.14, slightly higher than the 1.08 value obtained for the rGO aero. The difference in defect density can be attributed to the distinct synthesis and post-processing methods used to obtain the final material form. The aerogel formation process, combined with more gentler freeze-drying, effectively preserves the interconnected 3D network of rGO sheets, resulting in a more structurally intact material. In comparison, rGO HT in a powder form, dried at 60 °C under vacuum experienced stronger mechanical stress, which led to the creation of new defects and more intensive rGO sheet aggregation [37]. Therefore, despite identical GO precursor, the physical form of rGO directly influences the structural integrity of the material.

We have performed TGA to gain in-depth insight on the thermal stability of the reduced graphene oxides. Figure 3c shows that the rGO aero and rGO HT samples at 200 °C exhibit weight losses of 5.9and 12.2%, respectively. These losses are attributed to the evaporation of adsorbed water and the thermal decomposition of unstable oxygen-containing functional groups such as carboxyl groups, which are predominantly located at the edges of the rGO sheets and decompose to form CO, CO₂, and H₂O. As the temperature rises, more stable oxygen functionalities (e.g. hydroxyl, carbonyl and epoxy groups) begin to decompose, releasing additional CO and CO₂ [38]. Upon further temperature increase, the mass loss becomes more obvious. At 800 °C, the rGO aero and rGO HT retain 60.0 and 46.8% of their initial mass, respectively. This clearly indicates that the aerogel-based rGO demonstrates superior thermal stability compared to its vacuum-dried counterpart. The observed resutls can be attributed to the efficient removal of water through freeze-drying, which preserves the porous structure and allows for gradual deoxygenation without collapse. Additionally, the network of interconnected graphene layers of the aerogel resists abrupt degradation or oxidation of the carbon matrix, resulting in a higher residual mass at elevated temperatures compared to the rGO HT [39].

XPS analysis was performed to comprehensively investigate the surface chemistry of the reduced graphene oxides. The XPS survey spectra are shown in Figure S1, revealing the presence of two elements: C and O resulting from the hydrothermal reduction of GO without the use of chemical reductants, further underscores the simplicity of the synthesis process. Detailed information on the binding energies and full widths at half maximum (FWHM) of the C1s and O1s peaks is summarized in Table S1. The surface elemental composition and the functional groups distribution of rGO aero and rGO HT are very similar (Table 2) which suggests that the synthesis method and resulting physical structure did not significantly affect the chemical structure of the rGO surface. The C1s high-resolution spectrum of the rGO aero (Fig. 4a) revealed six components, with a dominant peak corresponding to C–C bonds in a sp² hybridization (284.6 eV). This confirms the restoration of the graphene hexagonal structure during the hydrothermal treatment of GO [23]. The peak at 285.4 eV is attributed to carbon in the sp³ configuration. Three additional peaks correspond to oxygen functional groups present on the surface of the material: hydroxyl/epoxy (286.1 eV), carbonyl/quinone (287.8 eV), and carboxyl groups (288.8 eV). Satellite peaks for C–C bonds at 290.4 and 291.9 eV are also observed [40].

Bild vergrößern

The O1s XPS spectrum of rGO aero was fitted by four components (Fig. 4b). Two of them are attributed to oxygen functional groups, i.e. carbonyl/carboxyl groups (531.3 eV) and hydroxyl/epoxy groups (533.3 eV). The component at 535.8 eV is linked to the adsorption of water on the rGO aero surface [41, 42]. The presence of this peak may be explained by developed porosity of the aerogel material which easily adsorbs humidity from the surroundings. The last peak (538.3 eV) can be related to the chemisorbed oxygen [41]. The noticeable enhancement occurs through increased polarity and wettability of carbon materials in the aqueous electrolyte solution [20, 33] and certain oxygen groups contribute to reversible Faradaic reactions, such as the Q-HQ reaction. This pseudocapacitance effect becomes particularly significant in an acidic environment, where a proton originating from the electrolyte participates in the process [20].

The assessment of the electrochemical performance of the two-electrode devices with rGO-based electrodes started with CV measurements with varying scan rates from 1 to 100 mV s⁻¹. In the beginning, we have compared the symmetric supercapacitors in a neutral 1 M Na₂SO₄ electrolyte (Fig. 5). The potential window for the system with both rGO materials remarkably exceeds the theoretical potential of water decomposition (1.23 V), reaching up to 1.8 V for rGO HT and 2.0 V for the rGO aero in 1 M Na₂SO₄ (Fig. 5a and b). The CV-derived efficiencies (coulombic and energy) show that coulombic efficiency (CE) remains around 95% for the rGO aero-based devices across explored potential window, while the energy efficiency (EE) decreases only slightly by 2%, when extending the maximum operating voltage from 1.6 to 2.0 V (Fig. S2). Thus neutral electrolytes without the abundance of H⁺ or OH^– ions provide higher kinetic barriers for the hydrogen and oxygen evolution reactions. The use of the rGO aero allows the window to be extended by 0.2 V in respect to rGO HT, resulting in a higher capacitance value. The voltammogram of the rGO aero in a neutral electrolyte exhibited a quasi-rectangular shape, indicating the ideal double-layer capacitive behavior [43]. Given that constant current is employed in commercially available devices rather than a constant voltage rate, as in CV, GCD measurements offer a more accurate representation of real supercapacitor performance. Despite using a safe potential range assessed from the voltammograms, GCD for rGO HT displayed asymmetry, suggesting lower CE (Fig. 5c). This distortion is attributed to exceeding the electrolyte stability window. However, the GCD curve for the rGO aero (Fig. 5d), even at higher voltage, maintained a close triangular shape, indicating nearly ideal capacitive behavior with a high CE close to 100% [44]. The specific capacitance at a current density of 0.2 A g^–1 is higher for rGO aero than for rGO HT (206 vs 143 F g^–1). Although the materials exhibit comparable S_BET area and elemental composition, the improved electrochemical performance of the rGO aero can be explained with structural differences, as revealed by Raman spectroscopy analysis. In our studies, the rGO aero presents I_D/I_G ratio of 1.08, while for rGO HT the value is marked higher, of 1.14. Our findings are in direct agreement with recently published work by Liu et al.[45], which provides compelling evidence for the relationship between the structural disorder and the electrochemical performance in carbon-based supercapacitors. The study reported that nanoporous carbons with broader D bands and smaller I_D/I_G intensity ratios exhibit higher capacitance due to smaller graphene-like domains as confirmed by NMR spectroscopy studies. Therefore, we conclude that the enhanced charge storage in rGO aero is due to a beneficial structural disorder characterized by smaller graphitic domains, which provide a higher density of accessible active sites for ion storage. Electrochemical performance measurements revealed that rGO aero facilitates supercapacitor operation with better parameters, enabling higher energy density values, a crucial aspect in energy storage applications. Therefore, further research will explore the use of a graphene aerogel in other aqueous electrolytes.

Bild vergrößern

The application of 1 M H₂SO₄ and 1 M H₂SO₄ + 0.2 M HQ as electrolytes allowed a broad potential range of 0–1.4 V to be achieved exceeding the theoretical potential value for water decomposition (Fig. 6a and b). CV profile of the supercapacitor operating in 1 M H₂SO₄ shows distorted rectangular shape due to pseudocapacitive effects of oxygen functional groups, which have a prominent electrochemical response in acidic electrolyte. While the addition of the redox-active compound did not alter the operating potential window, it facilitated higher capacitance values. In the case of 1 M H₂SO₄ coupled with 0.2 M HQ electrolyte, a pair of broad redox peaks appeared, with maxima around 0.18 and 0.50 V corresponding to the oxidation of HQ to quinone (Q) and the reverse reaction—reduction of Q to HQ, respectively [46]. Figure 6b shows the galvanostatic charge–discharge curves of the supercapacitors examined in different electrolytes. The curve for the 1 M H₂SO₄ system shows a similar discharge time compared to 1 M Na₂SO₄. However, due to the broader window in a neutral electrolyte, the acidic electrolyte allows higher capacitance values to be obtained (258 vs. 206 F g⁻¹ for 1 M Na₂SO₄). The positive impact of the addition of HQ is evident, resulted in an increase in the charge–discharge cycle time by more than 60% compared to 1 M H₂SO₄. The capacitance value for the device working in 1 M H₂SO₄ + 0.2 M HQ electrolyte is the highest (288 F g⁻¹) due to the fact it is a sum of double-layer capacitance, Faradaic surface pseudocapacitance, and Faradaic redox reactions of electrochemically active species [46]. Therefore, the capacitance values of the rGO aero measured in 1 M Na₂SO₄ at low current densities (0.2–1 A g⁻¹) are significantly lower than those of the redox-active system (Fig. 6e). The comparison of capacitance values at current densities of 0.2 and 10 A g⁻¹ reveals that the incorporation of 0.2 M HQ significantly impairs the system’s ability to maintain high capacitance. Under the highest current load of 10 A g⁻¹, the system using 1 M H₂SO₄ demonstrated the greatest tolerance, retaining 28% of the capacitance recorded at the lowest current density, in contrast to the redox electrolyte-based supercapacitor, which retained only 1%. At higher current densities, in the HQ-enriched electrolyte, challenges related to the diffusion of HQ active species at the electrode–electrolyte interface resulted in lower capacitance values compared to a system based on an electrolyte without HQ [47]. The electrochemical kinetics of the rGO aero symmetric supercapacitors in acidic electrolytes were investigated through the relationship between peak current i_p and scan rate ν, as presented in Fig. 7a. The log–log analysis of i_p versus ν yielded b-values of 0.91 and 0.65 for cathodic processes in 1 M H₂SO₄ and 1 M H₂SO₄ + 0.2 M HQ, respectively. If the b-value is close to 1, it reflects a charge storage mechanism governed predominantly by capacitive processes. However, the b-value of 0.5 represents diffusion-controlled processes [48]. Therefore, we observe a significant shift from capacitive to more diffusion-controlled electrochemical processes of the rGO aero symmetric supercapacitor, when redox-active HQ is added into the electrolyte. This is further supported by the deconvolution of the current contribution to surface-controlled current and diffusion-controlled current using Dunn method [49]. For the rGO aero 1 M H₂SO₄ + 0.2 M HQ, at a scan rate of 5 mV s^–1, the device shows almost 85% of the diffusion-controlled processes, and even at 100 mV s⁻¹, these processes still contribute to 52.3% of the current response (Fig. 7b).

Bild vergrößern

Bild vergrößern

The investigation of rGO aero-based supercapacitors using EIS measurements is illustrated through the Nyquist plots shown in Fig. 7c and d. Additionally, the equivalent circuit fitted using Zfit software is included in the inset in Fig. 7d. Notably, the fitted electrical circuit for devices with an electrolyte lacking a redox-active additive reveals an equilibrium differential capacitance element [43]. The bulk resistance (R_s), determined as the intercept at the real part (Z′), is referred to as a sum of the electrolyte resistance and current collector resistance [43]. The values of R_s are significantly influenced by the electrolyte used, being 0.65, 0.31, and 0.53 Ω for supercapacitors operating in neutral, acidic, and redox electrolytes, respectively (Fig. 6a). All supercapacitors exhibit a small semicircle in the high-frequency region (Fig. 7d), corresponding to the charge-transfer resistance (R_CT) caused by Faradaic reactions and double-layer capacitance on the graphene aerogel surface. The R_CT values, associated with ion transfer at the interfaces of electrode and electrolyte and the electron transfer between the current collector and the electrode are 0.78, 0.37, and 0.24 Ω for supercapacitors using 1 M Na₂SO_4, 1 M H₂SO₄, and 1 M H₂SO₄ + 0.2 M HQ electrolytes, respectively.

Despite the same anion (SO₄²⁻) in all electrolytes, the different cations (Na⁺ vs. H⁺) influence their electrical conductivity, resulting in higher conductivity for H⁺, explaining the lower resistance values for supercapacitors operating in sulfuric acid. Furthermore, the oxygen functional groups, abundantly present on the surface of the rGO aero, interact favorably with HQ, enabling the redox reaction and positively influencing system performance [20]. The slope of the straight line in the low-frequency region is related to the diffusion resistance in the electrode material. A more rapidly rising line toward the Z" axis indicates a low electronic resistance and better capacitive behavior [43] The calculated Warburg coefficient, representing the diffusion coefficient of ions, is 0.66 Ω s^−1/2 for the rGO aero supercapacitor with a neutral electrolyte. By contrast, it is 7.2 and 2.3 Ω s^−1/2 for setups with acidic and redox-active additive, respectively.

Further evaluations of the symmetric supercapacitors were conducted, focusing on calculated energy and power densities, as shown in the Ragone plot (Fig. 8a). Although the capacitance values are lower, the broad potential window enables the supercapacitor operating in a neutral electrolyte to achieve the highest energy density values. The best setup, employing 1 M Na₂SO₄, demonstrated impressive electrochemical performance with a high-energy density of 28.5 Wh kg⁻¹ at a power density of 0.154 kW kg⁻¹. In comparison, the devices using 1 M H₂SO₄ and 1 M H₂SO₄ + 0.2 M HQ exhibited energy density values of 17.5 and 19.5 Wh kg⁻¹ at power densities of 0.100 and 0.072 kW kg⁻¹, respectively. The observable decrease in stored energy for a device with a redox-active electrolyte above is related to the diminished capacitive capabilities of the setup at higher current regimes, at which the utilization of the faradaic redox reaction of Q-HQ species is significantly hindered. In addition to the high specific capacitance and thus high-energy density, the rGO aero–based devices exhibited outstanding stability during long-term cycling. Figure 8b presents the changes in the specific capacitance vs. cycle number for the assembled supercapacitors at a current density of 2 A g ⁻¹. The rGO aero in 1 M Na₂SO₄ operated within a potential window significantly exceeding the theoretical potential of water decomposition and maintained stability over thousands of cycles. Unfortunately, there was a gradual loss of charge storage capability, and the device reached 67% of its initial capacitance at the end of the experiment. By contrast, the supercapacitor operating in 1 M H₂SO₄ consistently retains almost 100% efficiency for 10 000 cycles. Furthermore, the device assembled with 1 M H₂SO_4, and 0.2 M HQ demonstrated outstanding stability, retained 98% of the initial capacity even after 12 000 charge–discharge cycles. Notably, a slight decrease in the specific capacitance value occurred at the initial cycles, followed by an increase in electrochemical capacitance and stabilization. The initial capacity fading observed during the first 1000 cycles can be attributed to the gradual adsorption and stabilization of the HQ species onto the carbon electrode surface. Such adsorption arises from electrostatic and π–π dispersion between the aromatic ring of HQ and the graphene layers, and hydrogen bonding with oxygen-containing surface functionalities. [20, 46]. The stability of the acidic electrolyte enhanced with the redox-active compound surpassed that of the neutral aqueous solution in terms of capacitive properties. Moreover, the combination of 1 M H₂SO₄ + 0.2 M HQ with the porous rGO aero emerged as a promising setup for high-energy and long-life supercapacitors. However, it is expedient to operate the supercapacitor at a current load not exceeding 1 A g⁻¹.

Bild vergrößern

The rGO aerogel-based devices exhibit exceptional energy density values and are among the highest of the symmetric aqueous supercapacitors reported previously in the literature (Table 3). The assembled devices operating at potentials beyond the theoretical water decomposition limit demonstrate impressive energy and power densities, with durability extending over several thousand charge/discharge cycles. The systems presented here offer a unique advantage in their simplicity and environmental compatibility, as they do not incorporate additional components into the active material during synthesis, making them a more sustainable alternative to the graphene materials with added enrichments or additives.