Modulating the supercapacitance of N-doped reduced graphene oxide synthesized using water-soluble inorganic precursors

Abstract

Controlling nitrogen functionalities in graphene are crucial for high-performance supercapacitors. In this study, we report the scalable synthesis of nitrogen-doped reduced graphene oxide (N-rGO) via a hydrothermal process using readily available inorganic precursors in an aqueous phase, enabling controlled manipulation of nitrogen functionalities. The optimized N-rGO (GO:aqueous NH₃ = 1:6) exhibits a areal capacitance of 0.4 F cm⁻² at a scan rate of 5 mV s⁻¹. In a symmetric two-electrode system, N-rGO delivers a high specific capacitance of 532 F g⁻¹, along with an energy density of 41 Wh kg⁻¹ and a power density of 375 W kg⁻¹ at 0.5 A g⁻¹, with a capacitance retention of ∼78% after 10,000 cycles at 1.96 V. These performance metrics are attributed to an optimal defect concentration combined with a high proportion of amino and pyrrolic nitrogen functionalities. Quantum capacitance calculations reveal an enhanced electronic density of states near the Fermi level, thereby improving charge storage performance.

Graphical Abstract

1 Introduction

Extensive research efforts have been dedicated to improving the efficiency and sustainability of energy storage systems, necessitating advancements in both energy and power density. With increasing global energy demand, the development of efficient energy storage technologies has become imperative. Batteries and supercapacitors represent two major classes of energy storage devices. Batteries typically exhibit high energy density but low power density, whereas supercapacitors demonstrate high power density but comparatively lower energy density. These limitations restrict their standalone applications as freelance energy storage devices (Zhi et al., 2008; Gogotsi and Simon, 2011). Various strategies have been implemented to enhance device performance, such as the design of layered porous structures (Yoon et al., 2000; Xing et al., 2006), particularly through the careful selection of suitable precursors. Currently, graphene and its derivatives have emerged as compelling electrode materials for achieving high-performance supercapacitors. The advantage of graphene is that it has a quite large surface area (over 2600 m²/g), good flexibility, increased electrical conductivity, mechanical strength, and acts as an ideal electrode material (Liu et al., 2010). These characteristics make it highly potential for an extensive range of applications, including transistors, sensors, energy conversion, and storage (Shao et al., 2010; Kamat, 2011; Pumera, 2011; Wang T. et al., 2014). In the current scenario, graphene production is still a tedious task on a large scale with an economical approach.

Extensive research has confirmed that graphene can be prepared in laboratory environments through various methods, including mechanical exfoliation of graphite (Allen et al., 2010; Yi and Shen, 2015), solvent exfoliation of graphite (Bourlinos et al., 2009; Khan et al., 2010), and growth on a metal substrate via chemical vapor deposition (CVD) (Obraztsov, 2009; Chen et al., 2011; Losurdo et al., 2011). As a substitute, several researchers prepared graphene oxide (GO) by easy chemical oxidation of graphite, which serves as a probable precursor for large-scale manufacture of graphene-related materials (Becerril et al., 2008; Huang et al., 2011). However, chemical oxidation treatments produce GO with many oxygen-containing functional groups. Due to these groups, graphene oxide becomes insulating, making it inappropriate for the purpose of the electrode material.

In contrast, the absence of a band gap in pristine graphene and the non-reactivity of its sp² hybridized carbon atoms render it suitable for such applications. To further deal with the limitations and to insert new properties via alien atoms, doping has attracted attention (Liu et al., 2011; Lv and Terrones, 2012; Wang et al., 2012). Recent research has demonstrated the successful doping of various heteroatoms, such as B, N, S, and P (Panchakarla et al., 2009; Paraknowitsch and Thomas, 2013), into the carbon lattice of graphene, thereby modifying its structural and chemical properties. Among several doping sources, nitrogen is the most abundant (along with its atomic radius being close to carbon) and thus most commonly used. Zhang and Xia (2011) demonstrated that doping of nitrogen could efficiently amend the spin density and charge distribution of adjacent carbons. Because of it, carbon domains are provoked and activated on the surface of N-rGO. Consequently, N-rGO illustrates promising activity toward electrocatalytic reactions. Furthermore, electronically rich N-dopant straightforwardly altered graphene’s electrical properties and makes N-rGO a probable applicant for sensors, electronics, and supercapacitors (Bose et al., 2012; Yavari and Koratkar, 2012; Hassan et al., 2013).

As a result, the enlargement of N-rGO unwraps new prospects for materials scientists to expand the utility of graphene. Alternatively, it is feasible to eliminate oxygen functional groups on GO sheets, incorporate the nitrogen atoms, and recover the conductivity by using suitable, economically favorable reduction treatments such as the hydrothermal/solvothermal method. Earlier researchers have done much work in this direction by utilizing different nitrogen precursors in various types of organic dispersive solvents (Long et al., 2010; Lai et al., 2011; Hasan et al., 2012; Lee et al., 2012; Lei et al., 2012; Zhao et al., 2012; Fan et al., 2014; Wang H. et al., 2014; Wang D.et al., 2014).

High-performance supercapacitors require the rational design of electrode materials with tunable electronic configuration and numerous active sites. Nitrogen-doped graphene is considered to be a promising material thanks to increased conductivity and pseudocapacitive nature of nitrogen dopants. However, the control of various forms of nitrogen dopants and their impact on capacitive performance is a challenging task that requires elaboration. Conventional techniques involving the use of nitrogen-containing organic precursors show insufficient tunability due to an uncontrollable decomposition process. To overcome this drawback, water-soluble inorganic nitrogen sources were applied in this study to synthesize reduced graphene oxide using a hydrothermal technique. It is more eco-friendly to use water as the solvent medium. Concentrations of various chemical compounds containing nitrogen were varied to obtain a tunable density of defects and nitrogen bonding states in reduced graphene oxide. Herein, the N-rGO samples were synthesized with different weight ratios of the nitrogen precursors and a fixed amount of GO (250 mg). It was shown that amino and pyrrolic nitrogen groups provide higher electrochemical performance than other nitrogen forms. The combination of experimental electrochemical studies with ab initio simulations allowed us to establish a structure–property relationship using quantum capacitance of doped carbon material.

2 Materials and methods

The syntheses of GO and N-rGO are provided below. Graphene oxide (GO) was synthesized using a modified version of the improved Hummers’ method, as reported in the literature (Yadav et al., 2018; Yadav et al., 2023). In brief, 3 g of graphite flakes were placed in a 500 mL beaker and mixed with a solution containing 360 mL of H₂SO₄ and 40 mL of H₃PO₄ under continuous stirring in an ice bath. Subsequently, 18 g of KMnO₄ was gradually introduced into the mixture while maintaining controlled temperature conditions. After completion of the reaction, the mixture was allowed to reach room temperature and was then thoroughly washed using deionized water, concentrated HCl, and ethanol to remove residual acids and impurities. The obtained material was collected by centrifugation and dried in an oven at 80 °C for 24 h.

Nitrogen-doped reduced graphene oxide (N-rGO) was prepared using various water-soluble inorganic nitrogen precursors, including NH₂OH, NH₄Cl, NH₄HCO₃, CH₃COONH₄, and aqueous/liquor NH₃ (liq. NH₃). For each precursor, three different GO-to-precursor weight ratios (1:4, 1:6, and 1:8) were employed. As a representative procedure, 1 g of NH₂OH was dissolved in 80 mL of deionized water, followed by the addition of 250 mg of GO. The resulting suspension was ultrasonicated for 15 min to ensure uniform dispersion. The homogeneous mixture was then transferred to a 100 mL Teflon-lined stainless-steel autoclave and heated at 150 °C for 24 h.

After naturally cooling to room temperature, the product was collected by centrifugation and repeatedly washed with deionized water and ethanol, with intermediate ultrasonication steps, to remove residual precursors and impurities. The purified material was dried overnight in a hot-air oven at 60 °C. The resulting sample was designated as N-rGO-4-NH₂OH.

Additional samples (N-rGO-6-NH₂OH and N-rGO-8-NH₂OH) were synthesized by varying the amount of NH₂OH while keeping the GO content constant (250 mg). The same procedure was followed for other nitrogen precursors, maintaining the corresponding weight ratios. Details of all synthesized samples are summarized in Table 1. For samples prepared using aqueous ammonia (aq. NH₃), the precursor quantity was controlled by volume (mL) rather than weight ratio.

Other details of the experiments and characterization are provided in Supplementary Material Section S1. For the three-electrode system, electrochemical measurements were conducted using a glassy carbon electrode (Supplementary Material Section S1). The specific details of sample preparation for the three-electrode system and the symmetric two-electrode device are provided in Supplementary Material Section S1.2. Cyclic voltammograms (CV), galvanostatic charge–discharge (GCD), and electrochemical impedance spectroscopy (EIS) experiments were performed for supercapacitor testing in a 0.5 M H₂SO₄ solution at room temperature (RT) without stirring.

3 Results and discussion

The scanning electron microscopy (SEM) micrographs for the synthesized samples are provided in Supplementary Figure S2. The SEM images of the N-rGO samples with various nitrogen precursors show sponge-like agglomerated lamellar structures. The X-ray diffraction (XRD) patterns in Figure 1, Panel I confirm the polycrystalline nature of the exfoliated GO sheets and N-rGO samples prepared from the different nitrogen precursors. Substantial peaks ascertained at 2θ ≈ 26.2°, resulting from the d₀₀₂ crystalline plane, illustrate the creation of reduced GO. As displayed in Figure 1, Panel 1:a–e, among all samples, the N-rGO-6 samples for each precursor have a sharp peak around 2θ ≈ 26.2°, which confirms the more crystalline nature of N-rGO-6 ratio samples in all cases compared to the N-rGO-4 and N-rGO-8 ratio samples.

FIGURE 1

Figure 1, Panel I:a–f illustrates the comparative XRD patterns of N-rGO-6 samples with different nitrogen precursors. From Figure 1, Panel I:a–f, we observed that N-rGO-6-CH₃COONH₄ and N-rGO-6-aq.NH₃ samples show a relatively sharper peak at 2θ = 26.2°, which illustrates the more graphitic nature and proper stacking of graphene sheets on each other compared to the other N-rGO-6 samples synthesized with the different nitrogen precursors. The broad peaks indicated the formation of amorphous-type carbon.

From the results shown in (ESI) Supplementary Table S1, all N-rGO samples with a 1:4 precursor ratio exhibit smaller crystallite sizes, likely due to incomplete exfoliation of graphene sheets. However, as nitrogen doping increases, crystallite size increases to N-rGO-6, which shows both the highest conductivity and the highest supercapacitive performance (as discussed later). After too much doping, the crystallite size declined for highly N-doped samples (N-rGO-8), indicating low conductivity and low supercapacitive performance among all the samples. From these results, we observed that a certain critical amount of nitrogen doping can enhance the conductivity and create the optimum amount of defects. For higher amounts of nitrogen doping, the number of defects further increases, thereby destroying the rGO sheets and resulting in a decline in conductivity and supercapacitive performance (discussed in Section 3.1, Electrochemical Study). An excess amount of nitrogen doping in the rGO sheets leads to increased defect density and amorphous-like characteristics (Yadav et al., 2022). This is attributed to the higher availability of nitrogen atoms. As a result, the full width at half maximum (FWHM) of the (002) peak becomes larger, indicating a decrease in the crystallite size for the N-rGO-8 samples. The (002) peak is fit with two sub-peaks correlating with the more defective (amorphous type) and less defective (graphitic type) carbon in the samples. The detailed study of the fitted XRD patterns is described in Supplementary Material Section 2.2.

3.1 Electrochemical study

Cyclic voltammograms, galvanostatic charge–discharge (GCD), and electrochemical impedance spectroscopy (EIS) were conducted using a three-electrode configuration in a 0.5 M H₂SO₄ electrolyte. The measurements were done within a potential range of 0–0.8 V, as depicted in the accompanying figure. Electrochemical impedance spectroscopy studies were performed over a frequency range of 200 kHz to 10 mHz, with a sine amplitude of 5 mV. Exact estimation of the material loading on the glassy carbon electrode (GCE, working area of 3 mm diameter) is “practically impossible.” Because the loading is probably in the μg range, the specific capacitance values can be highly overestimated. An accurate picture can be obtained with material deposition onto a larger area and with higher loading. Evaluating areal capacitance using the glassy carbon electrode is reasonable because the area is fixed. Therefore, the three-electrode measurement on a glassy carbon electrode is still useful for rapidly screening the better samples within a given series. The best sample is chosen for the device fabrication based on the results of the GCE for the areal capacitance.

Figure 2a illustrates the Cyclic voltammograms (CV) of N-rGO-6 samples, synthesized using various nitrogen sources, including different amounts of NH₂OH, NH₄Cl, NH₄HCO₃, CH₃COONH₄, and aq. NH₃. For additional CV and galvanostatic charge–discharge (GCD) curves of N-rGO-4, N-rGO-6, and N-rGO-8 samples synthesized with different precursors at extensive scan rates and current densities, please refer to (ESI) Supplementary Figures S4–S13. Across all these figures, it is evident that the voltammograms of N-rGO-4, N-rGO-6, and N-rGO-8 samples exhibit a near-rectangular shape with symmetrical current-potential characteristics at a scan rate of 5 mV/s. This suggests enhanced charge propagation within the electrode. The areal capacitance (C_Areal) values of the samples were determined from the CV curves based on the equation provided in Supplementary Material Section S2, and the obtained C_Areal values are presented in Table 2. The cyclic voltammograms of N-rGO-6 samples with all different nitrogen precursors at 5 mV/s scan rates are shown in Figure 2a, 2for N-rGO-4 and N-rGO-8 samples in Supplementary Figures S4–S13 in a potential window of 0–0.8 V (versus an Ag/AgCl reference). Among all N-rGO-4, -6, and -8 samples, the ones with a weight ratio of 1:6 (N-rGO-6 samples) show higher areal capacitance (at 5 mV/s) with the values of 0.148 F/cm², 0.235 F/cm², 0.230 F/cm², 0.359 F/cm², and 0.401 F/cm² for NH₂OH, NH₄Cl, NH₄HCO₃, CH₃COONH₄, and aq. NH₃, respectively (Supplementary Table S2). In Figure 2a, the capacitive regions are selected by performing CV experiments in the potential range where no oxidation–reduction reaction occurs (here, it is between 0 and 0.8 V). The CV curves show a rectangular contour at a 5 mV/s scan rate for all the samples, which shows their electric double-layer capacitor behavior. The current density of the N-rGO-6 samples (1:6 ratio) is the highest compared to N-rGO-4 (1:4 ratio) and N-rGO-8 (1:8 ratio) samples for all the different nitrogen precursors.

FIGURE 2

At lower scan rates, the ions (H⁺) in the H₂SO₄ electrolyte could diffuse into all the accessible sites, leading to absolute insertion reactions. Hence, the electrodes show almost ideal capacitive behavior. In contrast, with increasing scan rate, the diffused ions move toward only the outer surface of an electrode, and the productive communication between the ions and the electrode is significantly reduced, leading to deviations from the rectangular CV curve. In our case, the CV scans of samples show an almost rectangular contour at nominal scan rates displaying near-ideal capacitive performance. However, a slight deformation from a perfect rectangular contour was detected at high scan rates (Supplementary Figures S4–S13), which may be a feature because of escalating overpotential due to unfinished adsorption of ions onto the spots of the electrode material or from the ion transport between the electrolyte and the composite. The CV curves of the additional samples synthesized with different nitrogen precursors (N-rGO-4 and N-rGO-8 samples) also show the rectangular shape but demonstrate less current density (less I-V area) at 5 mV/s scan rates than N-rGO-6, which indicates the lowering of capacitance (see Supplementary Material). Among all the various nitrogen precursors, the N-rGO-6 sample synthesized using aq. NH₃ shows the highest areal capacitance.

Figure 2b illustrates the Nyquist plots for the N-rGO-6 samples synthesized with different nitrogen precursors, while the Nyquist plots for N-rGO-4 and N-rGO-8 can be found in Supplementary Figures S4–S13. The Nyquist plot can be separated into two regions: a high-frequency region and a low-frequency region. A semicircle is found in the high-frequency region, which is indicative of the electronic resistance of the electrode materials. A nearly straight line is observed in the low-frequency region, representing ideal capacitive behavior. Comparing the N-rGO-6 samples to the N-rGO-4 and N-rGO-8 samples, it is evident that the N-rGO-6 samples exhibit a more vertical line parallel to the imaginary axis in the low-frequency region, indicating superior capacitive behavior. Among the N-rGO-6 samples synthesized with different nitrogen precursors, the N-rGO-6-aq.NH₃ sample demonstrates the highest conductivity and displays a more pronounced vertical line in the low-frequency region than the other N-rGO-6 samples.

To investigate the factors contributing to the variation in the specific capacitance values of the synthesized samples, additional characterization techniques, including Raman spectroscopy and X-ray photoelectron spectroscopy (XPS), were employed. These spectroscopic analyses provide practical insights into the structural and chemical properties of the samples, elucidating their electrochemical performance. Raman spectroscopy was employed to analyze the presence of defects in the various carbonaceous phases of the N-rGO samples. Figure 1ii presents the Raman spectra of the N-rGO samples, allowing for the examination of the characteristic features associated with these carbon materials. The surface defects of the samples were evaluated based on the ratio of the graphitic (G) and disordered (D) peaks in the Raman spectra. Each spectrum exhibited two broad peaks located at approximately 1342 cm⁻¹ and 1583 cm⁻¹, which are characteristic of the carbonaceous materials being analyzed. The Raman spectra of the samples were further analyzed by fitting the peaks with Gaussian functions, which allowed for additional structural information to be extracted. The fitted spectra of all the samples can be found in Supplementary Figure S15, while the details of the sub-peaks are provided in Supplementary Table S3 (Yadav et al., 2021). The degree of graphitization is usually measured by the intensity ratio of I_G/I_D. Supplementary Table S4 shows that I_G/I_D values for all N-rGO-6 samples increased compared to N-rGO-4 and N-rGO-8 samples, which indicates that an optimum amount of nitrogen incorporation is required for better supercapacitive performance of N-rGO samples. The I_G/I_D values of all fitted spectra are given in Supplementary Table S4.

In addition, pure graphene oxide after reduction and simultaneous N-doping was characterized using X-ray photoelectron spectroscopy (XPS), which revealed that oxygen-containing functional groups were associated with all the sets of samples. A comparative study of intensities and areas under the peak indicated that not only did the nitrogen content vary from sample to sample but also the amounts of the different N-functional groups varied. These findings suggest that the synthesis conditions, viz., the nature of nitrogen source, synthesis temperature, pressure, and solvent, play important roles in introducing an optimum number of defects in the carbonaceous samples. The XPS results, therefore, confirm effective reduction and simultaneous N-doping in our samples. The comparative analyses of the %N content, N-environment, degree of reduction (C/O ratio), and disorder with the N-doped samples’ areal capacitance are presented in Table 2 and (ESI) Supplementary Table S2.

The percentage of different types of nitrogen environments is given in Table 2. Fitted wide spectra are provided in (ESI) Supplementary Figures S16,S17, and other elemental details are shown in (ESI) Supplementary Tables S5–S7. The oxygen-containing functional groups in graphene oxide (GO) were found to be responsible for reactions with nitrogen precursors, resulting in the formation of C-N bonds. The pyridinic group, which possesses a lone pair of electrons, imparts n-type doping behavior to the samples. In contrast, the pyrrolic groups display a more aromatic character, and graphitic nitrogen accounts for the augmentation of conductivity of the sample (Lee et al., 2013; Yadav et al., 2022). N-rGO-6-aq.NH₃ has a higher combined amount of amino, pyrrolic, and graphitic nitrogen, which increases the sample’s conductivity and capacitance. In contrast, the percentage of graphitic nitrogen is significantly less for the N-rGO-6-NH₂OH sample. Due to this, the sample was less conductive and showed substantially less capacitance. The results illustrate that the presence of higher amounts of amino, pyrrolic, and graphitic nitrogen in N-rGO-6-aq.NH₃ evidently boosts its capacitive behavior. The explanation may be ascribed to nitrogen’s high electronegativity, which helps to generate dipoles on the plane of reduced graphene oxide, resulting in superior affinity to pull in solvated charged species to the surface. From (ESI) and Table 2, Supplementary Table S3, it is notable that N-rGO-6-aq.NH₃ has the greatest I_G/I_D ratio, the highest predicted charge carrier concentration, and the highest combined proportion of amino, pyrrolic, and graphitic nitrogen among all the samples.

To examine the role of nitrogen doping in N-rGO-6-NH₄Cl samples (low % of N) and N-rGO-6-aq.NH₃ (moderate % of N) samples, the specific surface area, pore structure, and N₂ adsorption/desorption results are shown in Supplementary Figure S18 and (ESI) Supplementary Table S9, respectively. Both samples exhibit a typical IV-type plot characteristic of mesoporous characteristics (Yang et al., 2017). The specific surface areas of N-rGO-6-aq.NH₃ samples calculated by the BET method were the highest value (341 m² g⁻¹) because they maintain the ordered layered-like structure. On the other hand, the decrease in the specific surface area is much more severe after less nitrogen doping. Thus, the N-rGO-6-NH₄Cl sample illustrates a lower specific surface area (91 m² g⁻¹), which is most probably provoked by the low nitrogen doping effect (Yadav et al., 2022).

We have compared our results with existing literature in (ESI) Supplementary Table S10. Our sample shows the largest areal capacitance value, which indicates its superiority over other synthesis procedures and C_Areal values. By choosing a very simple eco-friendly and non-precious method, the procedure has the capability to scale to an industrial scale.

3.1.1 Device fabrication

From the earlier measurements, we found that the N-rGO-6-aq.NH₃ sample exhibits the highest areal capacitance in the three-electrode system, so, for a symmetrical two-electrode device system, we chose the N-rGO-6-aq.NH₃ sample. A CV curve from a 5 mV/s scan rate for the device is shown in Figure 3a. The fish-eye-shaped curve reveals some current leakage in the device. Figure 3b presents a galvanostatic cycling with potential limitation (GCPL) plot at 0.5 A/g current density for 10 cycles (lasting approximately 5905 s) using the symmetrical two-electrode device in a Swagelok cell, highlighting a quite good cyclic performance and high specific capacitance of 532 F/g. The device was then run for 10,000 charge–discharge cycles at a very high current density of 10 A/g (Figure 3c). The device’s cyclic performance was remarkably good (considering the rigorous electrochemical stress at 10 A/g), with capacitive retention of 77.4% and excellent Coulombic efficiency (97.7%) after 10,000 cycles. The inset figure shows the GCPL curves at a current density of 10 A/g for the first and the 10,000th cycles. The Ragone plot (Figure 3d) depicts an energy density of 41.5 Wh/kg and a power density of 375 W/kg at 0.5 A/g current density. Power density increases with current density, whereas the corresponding energy density decreases. This trend is consistent with typical supercapacitor behavior, in which rapid charge–discharge processes favor power delivery at the expense of energy storage. As shown in Figure 3e, the device was used to light a red LED, generating a potential difference of 1.96 V, showing its practicality.

FIGURE 3

3.2 Theoretical study

As shown in Table 2, the N-rGO-6 samples synthesized from the NH₂OH, NH₄Cl, NH₄HCO₃, CH₃COONH₄, and aq. NH₃ precursors have different concentrations of N-environments/functional groups, viz., pyridinic-N, amino-N, pyrrolic-N, graphitic-N, and oxidized-N. To understand the effect of these individual N-environments/functional groups in enhancing the specific capacitance of rGO sheets, we performed first-principles simulations using the density functional theory (DFT) framework (QuantumATK, 2023). Details of the methods used within the DFT framework are presented in Supplementary Material Section S1.4.

Figure 4 depicts the individual structures of various N-dopants/functional groups present in our synthesized N-rGO samples: pyridinic-N, amino-N, pyrrolic-N, graphitic-N, and oxidized-N. All carbon atoms of pristine graphene (Figure 4a) are in plane with each other, indicating their sp² hybridization, whereas in all other structures (Figure 4b–f), the carbon atoms bonded with the epoxy (-O-) group appear to be partially plucked out of the plane, signifying their hybridization change from sp² to sp³. A similar change of hybridization can be observed for the carbon atoms bonded with the amino group (-NH₂) and oxidized nitrogen (-NO) group. However, the N-atoms always remained in the plane with the carbon atoms in all rGO structures, portraying the sp² hybridization of the rGO.

FIGURE 4

The electronic nature of pristine graphene, graphitic-N in rGO, pyridinic-N in rGO, and pyrrolic-N in rGO has been previously reported by Saquib Khan et al. (2023). In addition, the amino-N in rGO and oxidized-N in rGO extracted in this work and depicted in Figure 5 show metallic and electronic properties. The electronic behavior of rGO sheets in the presence of individual N-dopants/functional groups is listed in Table 3. In general, metallic-natured materials are a good choice for the design of supercapacitor electrodes over semiconducting materials. Thus, Table 3 confirms that the incorporation of N-environments/functional groups into rGO sheets is good for the latter due to the induced metallicity. Earlier researchers have shown that a combination of pyrrolic, pyridinic, and graphitic-N plays a substantial role in increasing the conductivity of the doped graphene (Paek et al., 2013; Zhu et al., 2016). In order to evaluate the role played by these individual N-environments/functional groups in enhancing the specific capacitance of rGO, the quantum capacitances of these sheets have been extracted.

FIGURE 5

The internal capacitance exhibited by the rGO electrodes owing to their limited DOS at the Fermi level is called quantum capacitance (C_Q) and plays a significant role in deciding the specific capacitance of a supercapacitor device. In this work, the C_Q is estimated from the DOS profiles of N-doped/functionalized rGO sheets using The MathWorks Inc. (2026).

The expression of differential quantum capacitance of the supercapacitor electrode with synchronization correction for DOS and quantum capacitance is given by Equation 1 (Zhan et al., 2015; Pandey et al., 2021):

The quantum capacitance (C_Q) presented in this manuscript is given by Equation 2:

The computed quantum capacitances of the amino-N in rGO and oxidized-N in rGO are depicted in Figure 6, while the peak quantum capacitances offered by all the N-functional groups in rGO are listed in Table 3. The accuracy of the expressions and methods used for quantum capacitance extraction is validated in our earlier work (Srivastava and SanthiBhushan, 2018), in which the quantum capacitance of pristine graphene is found to align with the experimental report. From Table 3, the N-doped/functionalized rGO sheets exhibit peak quantum capacitances of 55.6 μF/cm² at 0 V for graphitic-N, 75 μF/cm² at 0.11 V for pyridinic-N, 91.2 μF/cm² at −0.66 V for pyrrolic-N, 187.7 μF/cm² at 0.02 V for amino-N, and 80.5 μF/cm² at −0.08 V for oxidized-N. The impressive peak quantum capacitance of the amino-N can be attributed to the large DOS peak available at the Fermi level in Figure 5. The peak quantum capacitance follows the pattern amino-N > pyrrolic-N > oxidized-N > pyridinic-N > graphitic-N. The amino-N offers the highest peak quantum capacitance, and pyrrolic-N offers the second-highest peak quantum capacitance, whereas graphitic-N offers the lowest peak quantum capacitance. Therefore, the results suggest that the rGO sample must have high amino-N and pyrrolic-N concentrations for the supercapacitor to offer high specific capacitance. This statement supports the data provided in Table 2, showing the high areal capacitance of N-rGO-6 samples obtained from CH₃COONH₄ and aq. NH₃ precursors, which have a relatively high combined amino + pyrrolic concentration of ∼67%, compared to other precursor-based samples. Thus, a higher combined amino + pyrrolic-N concentration may be the key to unlocking greater capacitance.

FIGURE 6

4 Conclusion

In the present study, a precursor engineering approach is proposed to modulate nitrogen functionalities in hydrothermally prepared N-rGO through water-soluble inorganic nitrogen precursors. Structural and spectroscopic investigations validate successful reduction and controllable functionalization of pyridinic, amine, pyrrolic, graphitic, and oxidized nitrogen atoms in N-rGO materials. Specifically, N-rGO prepared with a GO:aq. NH₃ mole ratio of 1:6 exhibits the highest areal capacitance of 0.4 F cm⁻², indicating the significance of precursor composition on the resultant properties. Computational investigation illustrates that pyrrolic and amine nitrogens play essential roles in enhancing electronic properties and quantum capacitance. A symmetric supercapacitor with optimized N-rGO delivers an impressive capacity of 532 F g⁻¹, with the energy and power densities of ∼41 Wh kg⁻¹ and 375 W kg⁻¹ at 0.5 A g⁻¹, respectively, and maintains ∼78% of its initial capacitance after 10,000 cycles.

References

• 1

  AllenM. J.TungV. C.KanerR. B. (2010). Honeycomb carbon: a review of graphene. Chem. Rev.110, 132–145. 10.1021/cr900070d

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  BecerrilH. A.MaoJ.LiuZ.StoltenbergR. M.BaoZ.ChenY. (2008). Evaluation of solution-processed reduced graphene oxide films as transparent conductors. ACS Nano2, 463–470. 10.1021/nn700375n

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  BoseS.KuilaT.MishraA. K.RajasekarR.KimN. H.LeeJ. H. (2012). Carbon-based nanostructured materials and their composites as supercapacitor electrodes. J. Mater. Chem.22, 767–784. 10.1039/c1jm14468e

  • CrossRef
  • Google Scholar

• 4

  BourlinosA. B.GeorgakilasV.ZborilR.SteriotiT. A.StubosA. K. (2009). Liquid-phase exfoliation of graphite towards solubilized graphenes. Small5 (5), 1841–1845. 10.1002/smll.200900242

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  ChenS.BrownL.LevendorfM.CaiW.JuS. Y.EdgeworthJ.et al (2011). Oxidation resistance of graphene-coated Cu and Cu/Ni alloy. ACS Nano5, 1321–1327. 10.1021/nn103028d

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  FanX.YuC.YangJ.LingZ.QiuJ. (2014). Hydrothermal synthesis and activation of graphene-incorporated nitrogen-rich carbon composite for high-performance supercapacitors. Carbon70, 130–141. 10.1016/j.carbon.2013.12.081

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  GogotsiY.SimonP. (2011). True performance metrics in electrochemical energy storage. Science334, 917–918. 10.1126/science.1213003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  HasanS. A.TsekouraE. K.SternhagenV.StrømmeM. (2012). Evolution of the composition and suspension performance of nitrogen-doped graphene. J. Phys. Chem. C116, 6530–6536. 10.1021/jp210474x

  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  HassanF. M.ChabotV.LiJ.KimB. K.Ricardez-SandovalL.YuA. (2013). Pyrrolic-structure enriched nitrogen doped graphene for highly efficient next generation supercapacitors. J. Mater. Chem. A1, 2904–2912. 10.1039/c2ta01064j

  • CrossRef
  • Google Scholar

• 10

  HuangX.LiS.HuangY.WuS.ZhouX.LiS.et al (2011). Synthesis of hexagonal close-packed gold nanostructures. Nat. Commun.2, 292. 10.1038/ncomms1291

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  KamatP. V. (2011). Graphene-based nanoassemblies for energy conversion. J. Phys. Chem. Lett.2, 242–251. 10.1021/jz101639v

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  KhanU.O’NeillA.LotyaM.DeS.ColemanJ. N. (2010). High-concentration solvent exfoliation of graphene. Small6, 864–871. 10.1002/smll.200902066

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 13

  LaiL.ChenL.ZhanD.SunL.LiuJ.LimS. H.et al (2011). One-step synthesis of NH2-graphene from in situ graphene-oxide reduction and its improved electrochemical properties. Carbon49, 3250–3257. 10.1016/j.carbon.2011.03.051

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  LeeJ. W.KoJ. M.KimJ. D. (2012). Hydrothermal preparation of nitrogen-doped graphene sheets via hexamethylenetetramine for application as supercapacitor electrodes. Electrochimica Acta85, 459–466. 10.1016/j.electacta.2012.08.070

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  LeeY. H.ChangK. H.HuC. C. (2013). Differentiate the pseudocapacitance and double-layer capacitance contributions for nitrogen-doped reduced graphene oxide in acidic and alkaline electrolytes. J. Power Sources227, 300–308. 10.1016/j.jpowsour.2012.11.026

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  LeiZ.LuL.ZhaoX. S. (2012). The electrocapacitive properties of graphene oxide reduced by urea. Energy Environ. Sci.5, 6391–6399. 10.1039/c1ee02478g

  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  LiuC.YuZ.NeffD.ZhamuA.JangB. Z. (2010). Graphene-based supercapacitor with an ultrahigh energy density. Nano Lett.10, 4863–4868. 10.1021/nl102661q

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 18

  LiuH.LiuY.ZhuD. (2011). Chemical doping of graphene. J. Mater. Chem.21, 3335–3345. 10.1039/c0jm02922j

  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  LongD.LiW.LingL.MiyawakiJ.MochidaI.YoonS. H. (2010). Preparation of nitrogen-doped graphene sheets by a combined chemical and hydrothermal reduction of graphene oxide. Langmuir26, 16096–16102. 10.1021/la102425a

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  LosurdoM.GiangregorioM. M.CapezzutoP.BrunoG. (2011). Graphene CVD growth on copper and nickel: role of hydrogen in kinetics and structure. Phys. Chem. Chem. Phys.13, 20836–20843. 10.1039/c1cp22347j

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  LvR.TerronesM. (2012). Towards new graphene materials: doped graphene sheets and nanoribbons. Mater. Lett.78, 209–218. 10.1016/j.matlet.2012.04.033

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  ObraztsovA. N. (2009). Chemical vapour deposition making graphene on a large scale. Nat. Nanotechnol.4, 212–213. 10.1038/nnano.2009.67

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  PaekE.PakA. J.KweonK. E.HwangG. S. (2013). On the origin of the enhanced supercapacitor performance of nitrogen-doped graphene. J. Phys. Chem. C117, 5610–5616. 10.1021/jp312490q

  • CrossRef
  • Google Scholar

• 24

  PanchakarlaL. S.SubrahmanyamK. S.SahaS. K.GovindarajA.KrishnamurthyH. R.WaghmareU. V.et al (2009). Synthesis, structure, and properties of boron- and nitrogen-doped graphene. Adv. Mater.21, 4726–4730. 10.1002/adma.200901285

  • CrossRef
  • Google Scholar

• 25

  PandeyS.KarakotiM.SuranaK.DhapolaP. S.SanthiBhushanB.GangulyS.et al (2021). Graphene nanosheets derived from plastic waste for the application of DSSCs and supercapacitors. Sci. Rep.11, 1–17. 10.1038/s41598-021-83483-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26

  ParaknowitschJ. P.ThomasA. (2013). Doping carbons beyond nitrogen: an overview of advanced heteroatom doped carbons with boron, sulphur and phosphorus for energy applications. Energy Environ. Sci.6, 2839–2855. 10.1039/c3ee41444b

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  PumeraM. (2011). Graphene-based nanomaterials for energy storage. Energy Environ. Sci.4, 668–674. 10.1039/c0ee00295j

  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  QuantumATK (2023). Synopsys Quantumwise A/S.

  • Google Scholar
  • View reference in article

• 29

  Saquib KhanM.SanthiBhushanB.Chandra BhamuK.GuK. S.Singh KushwahaH.SharmaA.et al (2023). Polymer hydrogel based quasi-solid-state sodium-ion supercapacitor with 2.5 V wide operating potential window and high energy density. Appl. Surf. Sci.607 (1–12), 154990. 10.1016/j.apsusc.2022.154990

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  ShaoY.WangJ.WuH.LiuJ.AksayI. A.LinY. (2010). Graphene based electrochemical sensors and biosensors: a review. Electroanalysis22, 1027–1036. 10.1002/elan.200900571

  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  SrivastavaA.SanthiBhushanB. (2018). Trade-off between quantum capacitance and thermodynamic stability of defected graphene: an implication for supercapacitor electrodes. Appl. Nanosci.8, 637–644. 10.1007/s13204-018-0643-x

  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  The MathWorks Inc. (2026). MATLAB version: 9.xx (R2026a). Natick, Massachusetts: The MathWorks Inc. Available online at: https://www.mathworks.com.

  • Google Scholar
  • View reference in article

• 33

  WangD.MinY.YuY.PengB. (2014). A general approach for fabrication of nitrogen-doped graphene sheets and its application in supercapacitors. J. Colloid Interface Sci.417, 270–277. 10.1016/j.jcis.2013.11.021

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  WangH.MaiyalaganT.WangX. (2012). Review on recent progress in nitrogen-doped graphene: synthesis, characterization, and its potential applications. ACS Catal.2, 781–794. 10.1021/cs200652y

  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  WangH.XieM.ThiaL.FisherA.WangX. (2014). Strategies on the design of nitrogen-doped graphene. J. Phys. Chem. Lett.5, 119–125. 10.1021/jz402416a

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  WangT.WangL.WuD.XiaW.ZhaoH.JiaD. (2014). Hydrothermal synthesis of nitrogen-doped graphene hydrogels using amino acids with different acidities as doping agents. J. Mater. Chem. A2, 8352–8361. 10.1039/c4ta00170b

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  XingW.QiaoS. Z.DingR. G.LiF.LuG. Q.YanZ. F.et al (2006). Superior electric double layer capacitors using ordered mesoporous carbons. Carbon44, 216–224. 10.1016/j.carbon.2005.07.029

  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  YadavA.KumarR.ChoudharyH. K.SahooB. (2018). Graphene-oxide coating for corrosion protection of iron particles in saline water. Carbon140, 477–487. 10.1016/j.carbon.2018.08.062

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  YadavA.KumarR.SahooB. (2020). Graphene oxide coatings on amino acid modified Fe surfaces for corrosion inhibition. ACS Appl. Nano Mater.3, 3540–3557. 10.1021/acsanm.0c00243

  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  YadavA.KumarR.PandeyU. P.SahooB. (2021). Role of oxygen functionalities of GO in corrosion protection of metallic Fe. Carbon173, 350–363. 10.1016/j.carbon.2020.11.029

  • CrossRef
  • Google Scholar
  • View reference in article

• 41

  YadavA.KumarR.SahooB. (2022). Exploring supercapacitance of solvothermally synthesized N-rGO sheet: role of N-doping and the insight mechanism. Phys. Chem. Chem. Phys.24, 1059–1071. 10.1039/d1cp03694g

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  YadavA.KumarR.KumarS.SahooB. (2023). Mechanistic insights into the roles of precursor content, synthesis time and dispersive solvent, in maximizing supercapacitance of N-rGO sheets. J. Alloys Compd.971, 172648. 10.1016/j.jallcom.2023.172648

  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  YangJ.ShinJ.ParkM.LeeG. H.Amedzo-AdoreM.KangY. M. (2017). The synergistic effect of nitrogen doping and para -phenylenediamine functionalization on the physicochemical properties of reduced graphene oxide for electric double layer supercapacitors in organic electrolytes. J. Mater. Chem. A5, 12426–12434. 10.1039/c7ta03194g

  • CrossRef
  • Google Scholar
  • View reference in article

• 44

  YavariF.KoratkarN. (2012). Graphene-based chemical sensors. J. Phys. Chem. Lett.3, 1746–1753. 10.1021/jz300358t

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 45

  YiM.ShenZ. (2015). A review on mechanical exfoliation for the scalable production of graphene. J. Mater. Chem. A3, 11700–11715. 10.1039/c5ta00252d

  • CrossRef
  • Google Scholar
  • View reference in article

• 46

  YoonS.LeeJ.HyeonT.OhS. M. (2000). Electric double-layer capacitor performance of a new mesoporous carbon. J. Electrochem. Soc.147, 2507–2512. 10.1149/1.1393561

  • CrossRef
  • Google Scholar
  • View reference in article

• 47

  ZhanC.NealJ.WuJ.JiangD. E. (2015). Quantum effects on the capacitance of graphene-based electrodes. J. Phys. Chem. C119, 22297–22303. 10.1021/acs.jpcc.5b05930

  • CrossRef
  • Google Scholar
  • View reference in article

• 48

  ZhangL.XiaZ. (2011). Mechanisms of oxygen reduction reaction on nitrogen-doped graphene for fuel cells. J. Phys. Chem. C115, 11170–11176. 10.1021/jp201991j

  • CrossRef
  • Google Scholar

• 49

  ZhaoY.HuC.HuY.ChengH.ShiG.QuL. (2012). A versatile, ultralight, nitrogen-doped graphene framework. Angew. Chem. - Int. Ed.51, 11371–11375. 10.1002/anie.201206554

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 50

  ZhiL.HuY. S.El HamaouiB.WangX.LieberwirthI.KolbU.et al (2008). Precursor-controlled formation of novel carbon/metal and carbon/metal oxide nanocomposites. Adv. Mater.20, 1727–1731. 10.1002/adma.200702654

  • CrossRef
  • Google Scholar

• 51

  ZhuJ.ChildressA. S.KarakayaM.DandeliyaS.SrivastavaA.LinY.et al (2016). Defect-engineered graphene for High-Energy- and high-power-density supercapacitor devices. Adv. Mater.28, 7185–7192. 10.1002/adma.201602028

  • Pubmed Abstract
  • CrossRef
  • Google Scholar