The Role of the Manganese Content on the Properties of Mn₃O₄ and Reduced Graphene Oxide Nanocomposites for Supercapacitor Electrodes

Abstract

Increasing the energy density and power of supercapacitors through hybrids of carbonaceous materials and metal oxides continues to be the subject of numerous research works. The correlation between specific capacitance and the properties of materials used as electrodes attracts great interest. In the present study, we investigated composites (GO/Mn₃O₄) prepared by the hydrothermal method with a variable ratio of GO/Mn₃O₄ and tested them as supercapacitor electrode materials in three- and two-electrode cells. The chemical characterization carried out by X-ray photoelectron spectroscopy and the adsorption techniques used allowed the determination of the surface carbon and oxygen content, as well as its textural properties. In this work, we analyzed the contribution of the double layer and the Faradaic reactions to the value of the final capacitance of the synthesized materials. Beyond empirically obtaining the electrochemical properties, these have been related to the physicochemical characteristics of the hybrids to help design materials with the best performance for supercapacitor electrodes.

1. Introduction

The high demand for energy to meet the needs of humanity, and the evidence of the damage that the combustion of fossil fuels causes to the environment [1], has tipped the balance towards clean forms of energy such as solar and wind power. The intermittency of this type of energy dictates the need to develop electrical energy storage devices, such as lithium-ion batteries (LiB) and supercapacitors [2,3]. The scientific community has proposed supercapacitors as an individual solution or in combination with other storage devices [3,4,5]. Supercapacitors have several advantages, including high power density, quick charge-discharge time, superior specific capacitance, low input resistance, long lifetime and are environmentally friendly [3,4,6]. Supercapacitors are the bridge between capacitors with high power density and batteries/fuel cells with high energy density [7,8,9]. Supercapacitors have greater energy density than capacitors but lower power density. Compared with conventional batteries, supercapacitors present an almost unlimited charge-discharge cycle life [10], have higher power density, but lower energy density [9].

The capacitance of a supercapacitor arises from two different basic mechanisms: the electrochemical double layer capacitance (EDLC), due to the separation of electrical charges at the Helmholtz layer, and the pseudo capacitance involving electrochemical Faradaic reactions (FR), in which a chemical reaction takes place in the electrode and the storage of electrical charge is carried out electrostatically, without interaction between the electrode and ions.

The search for new materials for supercapacitor electrodes is based on the modification of the precursors and on obtaining materials with high power, high energy density, fast charging and discharging and good tolerance to cycles [11]. Carbonaceous materials, such as biomass activated carbons [12,13,14], graphene [15,16,17,18,19], carbon cloths [20] and carbon nanotubes [21,22,23], have high conductivity, low cost and high specific surface area. In these materials, the predominant mechanism is EDLC; however, they have low capacitance [23]. Current research attempts to modify the structure and texture of these materials to, in turn, increase the specific capacitance. There have also been many attempts to combine them with transition metal oxides, which increases their operating voltage range and the specific capacitance [11,24]. Transition metal oxides present apparent advantages, including a variety of oxidation states [25,26,27], high specific capacitance and conductivity and low resistance, which make them very good candidates as electrodes for supercapacitors [28]. Manganese oxide, cobalt oxide, nickel oxide, zinc oxide, ruthenium oxide, among others, have been widely studied [24,29,30,31,32,33,34]. However, although the energy density provided by pseudocapacitive materials is always greater than the capacitance due to the double layer of EDLC provided by carbonaceous materials, they have lower stability due to the action of Faradaic reactions [19,22,35,36]. Hybrid supercapacitors aim to combine the advantages of both mechanisms by integrating carbon-based materials with metal oxide materials, thus incorporating physical and chemical charge storage mechanisms together in a single electrode [10,35,36]. Carbon-based materials provide a high surface area that increases the contact between the deposited pseudocapacitive materials and the electrolyte, and pseudocapacitive materials enhance the electrode capacitance through Faradaic reactions [37,38,39].

Manganese is an abundant, environmentally friendly, low-cost element with high theoretical specific capacitance; however, its performance is limited when a high-power supercapacitor is required, since it presents low electronic and ionic conductivity [11]. In an attempt to achieve more electroactive sites and faster ion transport, several studies have investigated Mn₃O₄ [40,41,42] with a porous structure, achieving a significant improvement in electrochemical performance. Hausmannite, or manganese oxide (Mn₃O₄), is a widely studied electrode material because of its fascinating features, such as high theoretical capacitance and variable oxidation states [43]. The improvement is even more significant when hybrids are prepared with graphene and graphene derivatives [42,44,45,46,47,48]. In the works found on hybrids based on reduced graphene oxide (rGO) and transition metal oxides (TMOs), it has been shown that the specific surface area and pore size distribution of the material play key roles in the capacitive behavior [11,44,49]. The distribution of TMO on a large surface substrate provides better electrical contact and improves the diffusion of the electrolyte to the active material. It has also been reported that the properties of the hybrids depend on the conditions of the formed graphene, defects, composition, etc. [44].

Many researchers have developed hybrids of Mn₃O₄ with great potential as energy storage with graphene derivatives, with excellent conductivity. Wang et al. [50] reported for the first time the preparation of graphene on the surface of Mn₃O₄ by CVD, obtaining a specific capacitance of 208.3 F g⁻¹ at 0.5 A g⁻¹. Jia et al. [51] prepared Mn₃O₄/rGO hybrids using a template method, reaching the material 561 F g⁻¹ at a current density of 1 A g⁻¹. However, due to the complex methods of synthesis and the difficulty in controlling the structure of the material, it remains a challenge to develop a simple method to prepare Mn₃O₄/rGO with control on the electrochemical properties.

Finding a good material involves understanding the delicate balance between the two mechanisms that can participate in the charge accumulation process, EDLC and Faradaic reactions, and designing the best possible material requires knowing the role that each component plays in each of these mechanisms. As far as we know, there have been no systematic studies of this type in the literature. With this objective in mind, in this work, we have prepared hybrids with different proportions of Mn₃O₄ nanoparticles decorating the surface of reduced graphene oxide using a simple hydrothermal method. The hybrids were characterized by Raman and XPS spectroscopies, XRD and gas adsorption of N₂ and CO₂. The variable amounts of Mn₃O₄ lead to hybrids with different electrochemical behavior tested by cyclic voltammetry and galvanostatic charge-discharge measurements. This study allowed us to understand the role that the two components of the hybrids play in the charge storage mechanisms and the contribution of each to the specific capacitance. This will help in designing reduced graphene oxide/Mn₃O₄ hybrid materials with the best electrochemical performance for supercapacitor electrodes.

2. Materials and Methods

2.1. Materials and Reagents

Graphene oxide was synthesized by the oxidation of natural graphite flakes (CAS 7782-42-5) from Sigma Aldrich (Madrid, Spain). The reagents used for graphite oxidation and for the preparation of hybrids—NaNO₃ (99%), H₂SO₄ (98% w), KMnO₄ (>99%), H₂O₂ (30% V) and manganese acetate Mn(CH₃COO)₂·4H₂O (>99%)—were purchased from Sigma Aldrich (Madrid, Spain). NaOH (>97%) was supplied by Panreac Química SLU (Barcelona, Spain). For electrochemical characterization, the reagents used were poly(vinylidene) fluoride (PVDF) from Sigma Aldrich, tetraethylammonium tetrafluoroborate Et₄NBF₄ from Alfa Aesar and acetonitrile (CAN) HiPerSolv CHROMANORM^® purchased from VWR. All reagents were used without purification.

We used ultra-pure water from a RiOs and Milli-Q combined system from Millipore. Carbon dioxide (CO₂), nitrogen (N₂) and helium (He) were supplied by L’Air Liquide, Madrid, Spain. The minimum purity was 99.999%.

2.2. Synthesis of Graphene Oxide, Metal Oxides and Graphene Oxide/Metal Oxide Hybrids

Graphene oxide (GO) was obtained through the oxidation procedure reported by Hummers and modified by our group [52,53,54,55]. The preparation of the graphene oxide/manganese oxide hybrids was carried out using acetate of manganese. An aqueous solution of NaOH (0.1 M) was added dropwise to a solution of Mn(CH₃COO)₂·4H₂O (5 mg/mL) until the pH value was approximately 12. To the colloidal solution obtained, a certain volume of the aqueous dispersion of GO (1 mg/mL) was added. The mixture was diluted to 30 mL with ultra-pure water and stirred for 6 h. Subsequently, the mixture was transferred to a 110 mL stainless steel Teflon-lined autoclave and heated from room temperature to 180 °C at 2 °C/min and then held at 180 °C for 12 h [56]. After this time, a black solid in the shape of a cylinder was obtained, which was collected by filtration after being cooled to room temperature, washed with ethanol three times and dried at 60 °C for 12 h in an oven. The graphene oxide/Mn(CH₃COO)₂·4H₂O weight ratio was modified with the volume of the starting manganese acetate solution, Mn(CH₃COO)₂·4H₂O (5 mg/mL), obtaining samples with weight ratios (GO/Mn(CH₃COO)₂·4H₂O) of (20/80), (50/50), and (80/20). Said samples were named GOMn14, GOMn11 and GOMn41, respectively. It is accepted that the synthesis reaction takes place in three stages [57]: the formation of manganese hydroxide in a basic medium; the formation of a coordination complex with the material carbonaceous, in this case with graphene oxide, which would help release Mn²⁺; and in the last stage, a partial oxidation of Mn²⁺ into Mn⁺³ would occur due to the limited existence of oxygen in the autoclave under the reaction conditions.

Since the hydrothermal process leads to reduced graphene oxide [58], for comparative purposes, graphene oxide was subjected to the same hydrothermal process used for manufacturing the hybrid materials. The reduced GO obtained was called rGO.

2.3. Structural, Chemical and Textural Characterization

X-ray photoelectron spectra (XPS) of powder samples were recorded in a PHI Versa Probe II (Physical Electronics, Chanhassen, MN, USA) equipped with an excitation source of Al Kα (1486.6 eV) at 25 W and a 1.3V and 20.0 μA neutralizer. The high-resolution spectra were recorded working at 29.35 eV analyzer pass energy.

Raman spectra were recorded at room temperature with a micro-Raman spectrometer LabRAM HR Evolution (Horiba Jobin-Yvon, Palaiseau, France). To obtain the Raman spectra of each material, solids are placed onto silicon wafers. Materials were irradiated with 532 nm light from a solid-state laser and a 100× objective (laser spot size 1 μm²). The laser excitation power was kept below 1 mW to avoid laser-induced heating. Calibration was performed by checking the Rayleigh band and Si band at 0 and 520.7 cm⁻¹, respectively. The area was scanned with a spatial resolution of approximately 0.5 μm, and the acquisition time was 3 s at each point. Each Raman spectrum was recorded at least in five different regions of the samples, and the spectra shown in figures are the average of all measurements. The diffraction grating used was 1800 gr mm⁻¹, and the spectral resolution with this configuration was 2 cm⁻¹.

Powder XRD patterns were recorded in a Bruker D8 Advance powder diffractometer using CuKα1,2 radiation (λ = 1.54050 Å) between 5° and 80° (2θ) with a step size of 0.05° and a step time of 2.6 s. The tube operated at 40 kV and 30 mA.

The porous texture of the samples was analyzed from physical adsorption isotherms of N₂ at 77 K and of CO₂ at 273 K, measured in an automatic Micromeritics ASAP 2010 volumetric adsorption apparatus. These isotherms were used to obtain the specific surface area S_BET(N₂), calculated by applying the BET equation to the N₂ adsorption isotherm. Micropore volume was determined from applying the Dubinin–Raduskhevic equation to the adsorption isotherms of CO₂, V_MP(CO₂), at relative pressures under 0.01 [59,60]. The adsorption of the two gases provides complementary information. Whereas N₂ is adsorbed into micropores larger than 0.7 nm, the adsorption of CO₂ at higher temperature takes place in smaller micropores (<0.7 nm).

2.4. Electrochemical Characterization

The electrochemical performance of the hybrid materials was studied in a three-electrode cell using an Arbin instruments BT-G multichannel potentiostat. For the preparation of the electrodes, synthesized hybrid materials (80 wt.%) were mixed with carbon black (10 wt. %) as the conductive material and PVDF binder (10 wt.%). After grinding in an agate mortar, ethanol was added to form a paste. Single-sided coated electrodes were made by depositing the paste on Ni foam over a 1 cm² area. The electrodes were dried at 90 °C overnight. Ag/AgCl electrodes and Pt wire were used as the reference and counter electrodes, respectively. The electrolyte employed was a 1 M KOH aqueous solution. Samples were also tested in a two-electrode cell with symmetric configuration, using Et₄NBF₄/CAN as the non-aqueous electrolyte.

3. Results

3.1. Structural, Chemical and Textural Characterization of Hybrid Materials

3.1.1. XPS, Raman Spectroscopy and Gases Adsorption

X-ray photoelectron spectroscopy (XPS) was used to evaluate the surface composition of the nanocomposites. Figure 1a shows the survey spectra of the different materials.

The presence of C, O and Mn atoms was revealed by the survey spectra plotted in Figure 1a and Figure S1 of the Supplementary Materials. Thus, bands at 285 eV and 530 eV, assigned to C1s and O1s, appear in all materials, while the two peaks centered at 642 and 653 eV, consistent with the binding energies of Mn 2p^3/2 and Mn 2p^1/2, respectively, appeared in the hybrids, and its intensity increased with the percentage of Mn used in the synthesis. The presence of these two bands in the survey spectrum indicates the presence of Mn₃O₄, which perfectly matches the previously reported values for hausmannite [61,62]. Hausmannite contains both manganous and manganic ions [63,64]. Therefore, if the difference in the binding energies of manganous and manganic ions in hausmannite is large enough, the XPS spectrum will have a larger doublet separation of the Mn 2p1/2 and Mn 2p3/2 spin-orbit levels than the simple oxides [61,62]. The XPS spectra of Mn 2p are analyzed in Figure S2 of the Supplementary Materials.

The atomic percentage of each element and the O/C ratio are presented in

Table 1. Atomic composition determined by XPS. Textural characterization by N₂ and CO₂ adsorption.

Materials	C1s Atomic (%)	O1s Atomic (%)	Mn2p Atomic (%)	O/C	SBET (m2/g)	VMP (cm3/g)
rGO	88.4 ± 3	11.5 ± 0.9	---	0.13	184	0.06
GOMn41	86.2 ± 4	13.2 ± 0.6	0.50 ± 0.03	0.15	346	0.10
GOMn11	68.8 ± 3	23.1 ± 1	8.0 ± 0.4	0.33	214	0.06
GOMn14	46.4 ± 2	35.5 ± 2	18.0 ± 0.9	0.77	125	0.04

. The comparison of the O/C ratio of rGO with the O/C ratio of the initial GO suspension, as previously published by our group (O/C _GOSA =0.52) [54], shows that the hydrothermal process produced a significant reduction of GO, since its O/C ratio decreased from 0.52 to 0.13 after hydrothermal treatment.

The atomic percentage of Mn obtained by XPS increases as the Mn content used in the synthesis of the hybrid increases.

Specific surface area calculated from the N₂ adsorption and BET model (S_BET) and the micropore volume (VMP) obtained from CO₂ adsorption isotherms and the Dubinin−Radushkevich model are also presented in Table 1.

To check if the graphene network was retained after the hydrothermal treatment, micro-Raman spectra of rGO and the hybrid materials were recorded and plotted in Figure 1b. As can be seen in Figure 1b, the Raman spectra of all materials present two well-defined bands: D-band, centered around 1350 cm⁻¹, and G-band, centered close to 1585 cm⁻¹, characteristic of disordered graphitic materials such as graphene oxide and reduced graphene oxide [53,54,65]. The G band is an allowed band by the selection rules and is assigned to the vibration of the sp² carbon lattice. The D band is activated by the presence of defects (holes or functional groups) in the flakes of graphene networks.

The presence of these bands in the Raman spectra means that the disordered graphene network is present in the hybrids prepared by the hydrothermal method. In addition, a third band centered at 668 cm⁻¹ appears in the spectra of the hybrid materials GOMn11 and GOMn14. This band was previously reported [56,66] and assigned to the stretching vibrations of the metal-oxygen bonds in Mn₃O₄. This demonstrates the formation of Mn₃O₄ in the hybrid materials GOMn11 and GOMn14. As expected, the intensity of this band increases with the proportion of Mn in the hybrid material.

3.1.2. XRD and SEM Measurements

To analyze the evolution of the crystallinity and the morphology with the hybridization process, we recorded the X-ray diffractograms and SEM images of the different materials. XRD diffractograms are plotted in Figure 2a,b, and SEM images are in Figure 2c–f, respectively.

It is interesting to note that the diffractogram of the hybrid with the highest percentage of rGO (GOMn41), Figure 2a, is dominated by two broad peaks centered at 24.69° and 43.48°, which can be attributed to diffractions corresponding to the (002) and (100) of graphene oxide (rGO) [58]; however, the diffraction peaks corresponding to manganese oxide are not observed.

On the contrary, the diffractograms corresponding to the hybrids with a high proportion of Mn, GOMn11 and GOMn14, included in Figure 2b, are constituted by narrow peaks centered at 19.05°, 29.50°, 31.16°, 32.60°, 36.10° and 44.55°, which can be assigned to the diffraction of the planes (101), (112), (103), (200), (211) and (220) of Mn₃O₄ (JCPDS No. 24-0734) [56,61].

To obtain information of the morphology of these hybrids, SEM images of the solids were taken, shown in Figure 2c–f. As can be seen in Figure 2c,d, the samples with a high content of rGO show disordered and wrinkled platelets. These results are compatible with the diffractograms shown in Figure 2a, which are mainly dominated by the peaks corresponding to rGO. In the hybrids with a high percentage of Mn₃O₄, Figure 2e,f, corresponding to GOMn11 and GOMn14, respectively, shows the presence of nanoparticles with diameters between 35 and 70 nm. This size is consistent with Mn₃O₄ nanoparticles reported by other authors [43,67]. In this matrix of nanoparticles, dispersed graphene sheets can be observed forming a non-crystalline phase, which would explain the background band observed in the diffractograms of Figure 2b.

3.2. Electrochemical Performances

3.2.1. Cyclic Voltammetry

Cyclic voltammetry (CV) was employed to evaluate the electrochemical properties of the electrodes prepared with the hybrid materials using a three-electrode system. We carried out CV measurements at 5, 10, 20 and 50 mV/s. Figure 3 compares the CV profiles obtained for rGO, GOMn41, GOMn11 and GOMn14 at scan rates of 5 and 50 mV s⁻¹, as references for low and high scan rates within the potential range of −1.5 to 0.2 V. The CV curve for rGO at 5 mVs⁻¹ (Figure 3a) exhibits a pseudo-rectangular shape without any distinct Faradaic peak in this potential window. Typically, a rectangular CV shape is characteristic of carbonaceous materials that store charge through the formation of the electric double layer at the electrode–electrolyte interface [68]. However, a rounded shape was detected in the cycle, suggesting some pseudocapacitive contribution due to the oxygenated functional groups on rGO [68]. On the other hand, the CV curves for GO/Mn₃O₄ hybrids deviate from the rectangular shape, exhibiting slight hump peaks at low potentials, a signature of pseudocapacitive behavior of the electrode materials.

Accordingly, it is possible to conclude that these hybrid materials exhibit both pseudocapacitive behavior and electric double layer capacitor (EDLC) behavior. At 50 mV s⁻¹, in Figure 3b, the peaks observed at low scan rate, assigned to Faradaic reactions, disappear, except for GOMn14 with the highest Mn content, in whose curve the deviation from the rectangular shape is slightly appreciated. At low scan rates (5 mVs⁻¹), there is ample time for redox reactions, while at high scan rates, the diffusion of reagents through the pores of the material hinders the occurrence of redox reactions. Finally, when the Mn₃O₄ percentage increases, the CV cycles become more symmetric, indicating improved electrochemical behavior and charge store capabilities of the hybrid materials.

3.2.2. Galvanostatic Charge Discharge GCD

The energy density, power density and specific capacitance of the hybrid materials were evaluated using the galvanostatic charge-discharge (GCD) technique. Figure 4a displays the GCD curves obtained for rGO, GOMn41, GOMn11 and GOMn14 at a current density of 1 A g⁻¹ performed between −1 and 0.2 V (vs. Ag/AgCl) in a 1 M KOH solution. The GCD curves are not perfectly linear, which may be related to the coexistence of two mechanisms: the electric double layer and pseudocapacitive mechanisms. As can be seen in Figure 4a, the charge-discharge time is clearly longer for the GOMn41 hybrid than for rGO, and shorter in the case of the GOMn14 hybrid.

The specific capacitances (SC), energy densities (E) and power densities (P) of the materials were calculated from GCD curves using Equations (1)–(3), respectively.

$$\mathrm{SC}\left({\mathrm{F}\mathrm{g}}^{-1}\right)=\frac{\mathrm{I}\Delta \mathrm{t}}{\Delta \mathrm{V}\mathrm{m}}$$

(1)

$$\mathrm{E}\left({\mathrm{W}\mathrm{h}\mathrm{kg}}^{-1}\right)=\frac{1}{2}\mathrm{}\frac{\mathrm{SC}{\left(\Delta \mathrm{V}\right)}^2}{3.6}$$

(2)

$$\mathrm{P}\left({\mathrm{W}\mathrm{kg}}^{-1}\right)=\mathrm{}\frac{3600\mathrm{E}}{\Delta \mathrm{t}}$$

(3)

where SC represents the specific capacitance (F g⁻¹), I (A) is current, ∆V (V) is the total potential of the voltage window, ∆t (s) is the discharge time and m (g) is the mass of the active electrode material.

The specific capacitance, energy and power density values calculated from data obtained at a current density of 1 A g⁻¹ are collected in

Table 2. Specific capacitance in KOH 1 M (SC), SC after 500 cycles and energy and power densities. Specific capacitance in Et₄NBF₄/CAN (two-electrode cell). SC after 500 cycles, energy and power densities.

	KOH 1 M (Three-Electrode Cell)				Et4NBF4/CAN (Two-Electrode Cell)
Materials	SC (F g−1)	E (W h kg−1)	P (W kg−1)	SC Retention after 500 Cycles (%)	SCEDLC (F g−1)
rGO	137 ± 27	27	736	73 ± 15	90 ± 18
GOMn41	173 ± 34	34	753	76 ± 15	66 ± 13
GOMn11	136 ± 26	26	761	74 ± 15	35 ± 7
GOMn14	89 ± 17	17	768	71 ± 14	17 ± 3

. The SC values calculated were 137, 173, 136 and 89 F g⁻¹ for rGO, GOMn41, GOMn11 and GOMn14, respectively.

The SC values obtained at different current density values are plotted in Figure 4b. The values decreased by increasing the current density from 1 to 5 A g⁻¹. This can be due to the limited interaction time between electrolyte ions and the inner regions of the electrode material at elevated current densities [69]; the ions have no opportunity to diffuse into the pores of the material [70].

The values obtained for energy and power densities have been collected in Table 2 and plotted in Figure 4c. The Ragone plot of rGO and GOMn₃O₄ hybrids (Figure 4c) provides the relationship between the energy and the power density. The cyclic stability test was performed using the galvanostatic charge-discharge technique at a constant density of 1 Ag⁻¹ in the potential range of 0.0 V to 0.2 V for 500 cycles. The specific capacitance retention after 500 charge-discharge cycles is also presented in Table 2 and shows that rGO/Mn₃O₄ composites have good cycle life as an electrode material for supercapacitors since they maintain at least 71% of the initial SC.

4. Discussion

We are interested in analyzing the correlation between the specific capacitance (SC) in KOH 1 M and the structural and textural properties. With this purpose in mind, in Figure 5a, the specific capacitance in a three-electrode cell, SC, is plotted against the specific surface area (SBET). To evaluate the contribution of the double-layer mechanism to the total specific capacitance, all materials were also tested in a two-electrode cell with a symmetrical configuration using Et₄NBF₄/CAN as a non-aqueous electrolyte (Table 2). We designated the values of SC thus obtained as SC_EDLC, since pseudocapacitive reactions do not occur in an organic medium. Figure 5b shows the variation of the specific capacitance in two-electrode cells, SC_EDLC, with the O/C ratio calculated from XPS. Finally, as we expected Mn to be responsible for the Faradaic reactions, SC_FR, estimated from SC_FR = SC − SC_EDLC, was plotted against the atomic percentage of Mn obtained from XPS measurements in Figure 5c.

The plots reveal interesting information. Figure 5a shows a linear correlation between specific capacitance and specific surface area. This behavior agrees very well with results found for carbon-based supercapacitors [71,72]. Accordingly, GOMn41, with the highest S_BET, shows the best performance as a supercapacitor electrode, while both rGO and GOMn11, which share similar S_BET values, exhibit comparable specific capacitance values of ~136 F g⁻¹. A similar trend is observed when correlating specific capacitance with the micropore volume determined through CO₂ adsorption, as seen in Table 1. This behavior has previously been reported for carbonaceous materials [72] and attributed to the existence of an electric double-layer mechanism. The authors considered that the inclusion of carbonaceous materials enhances the specific surface area of the electrodes, facilitating the ion contact and promoting the EDLC mechanism, which relies on ion adsorption [72].

Figure 5b represents the specific capacitance values obtained in a two-electrode cell with an Et₄NBF₄/CAN medium, SC_EDLC, against the O/C atomic ratio obtained by XPS (Table 1). As can be seen in Figure 5b, SC_EDLC shows a continuous decrease as the O/C ratio increases. Accordingly, rGO, which has the lowest O/C ratio, demonstrates the best electrochemical performance in this medium.

Although it is acknowledged that isolating the contribution of the electric double-layer mechanism from that of Faradaic reactions to the total specific capacitance is challenging, we have attempted a naive approach to separate both contributions by calculating the difference between the SC values obtained in KOH medium using a three-electrode cell (SC) and those obtained in an organic medium using a two-electrode cell (SC_EDLC). This is an attempt to distinguish the Faradaic mechanism and explore its relationship with certain structural properties. Although Faradaic reactions can be expected to be favored by Mn content, Figure 5c shows that the SC_FR of the hybrids is always greater than the specific capacitance of rGO; however, the SC_FR decreases when the percentage of Mn increases.

As mentioned above, the hybrid with the lowest percentage of Mn, GOMn41, manages to improve the specific capacitance of rGO from 137 to 173 F g⁻¹, increase its energy density from 27 to 34 W h kg⁻¹ and power from 736 to 753 W kg⁻¹. Likewise, GOMn41 has the highest capacitance retention after 500 cycles at 1 A g⁻¹. As mentioned, this effect may be due, on the one hand, to the higher proportion of carbon, which favors the EDLC effect, and on the other hand, to the increase in the accessible surface of the material, thanks to the presence of Mn₃O₄ nanoparticles that serve as spacers, avoiding the stacking of rGO sheets [73].

For hybrids with a higher proportion of Mn₃O₄ nanoparticles, the specific capacitances decrease from 137 F g⁻¹ for rGO and to 136 and 89 F g⁻¹ for GOMn11 and GOMn14, respectively, and the energy densities also decrease. The increase in Mn₃O₄ nanoparticles on the surface decreases the specific surface area (S_BET) due to the blocking of the pores, as observed in Table 1 and in the SEM images.

We can conclude that the presence of Mn nanoparticles favors the presence of Faradaic reactions and will increase the specific capacitance, to the extent that the specific surface area is not compromised.

5. Conclusions

In summary, we synthesized supercapacitor electrode materials using Mn(CH₃COO)₂ and simultaneously reduced graphene oxide (rGO) through a single-step hydrothermal process.

The GOMn41 hybrid, characterized by the lowest percentage of Mn, demonstrates significantly enhanced capacitive performance compared to graphene oxide treated with hydrothermal methods, rGO, and exhibits the highest capacitance retention after 500 cycles. This improvement is attributed to the unique structure achieved by combining rGO and Mn₃O₄ nanoparticles in a 4:1 weight ratio, resulting in a high specific surface area that enhances the initial specific capacitance of rGO. Hybrids with a higher proportion of Mn₃O₄ result in a less accessible structure probably due to pore blocking. The content of Mn₃O₄ nanoparticles favors the presence of Faradaic reactions and will increase the specific capacitance, if the specific surface area is not compromised. The results shed light on the preparation of hybrid materials derived from Mn₃O₄ and reduced graphene oxide for use as supercapacitor electrodes.