Those numbers matter because they were achieved without organic solvents, precious catalysts, or the uncontrollable decomposition chemistry that has plagued nitrogen-doped graphene synthesis for years. Most previous efforts relied on nitrogen-containing organic precursors such as urea, amino acids, or amines dissolved in organic dispersive solvents. When these compounds decompose under heat, they break apart in ways that are hard to predict, so the final material ends up with an unpredictable mix of nitrogen bonding states. The new work sidesteps that problem entirely by starting from ordinary inorganic compounds—hydroxylamine, ammonium chloride, ammonium bicarbonate, ammonium acetate, and aqueous ammonia—and letting water do the dispersing. The synthesis is a standard hydrothermal process: graphene oxide is mixed with the precursor in a Teflon-lined steel autoclave and heated to 150 degrees Celsius for 24 hours. That is a recipe any chemistry lab, and potentially any factory, could follow.

The crucial insight behind the study is that not all nitrogen atoms are created equal. When nitrogen is incorporated into the carbon lattice of reduced graphene oxide, it can sit in several distinct chemical environments. Pyridinic nitrogen occupies an edge site with a lone electron pair and lends n-type character to the material. Pyrrolic nitrogen bonds into five-membered aromatic rings and carries a strongly aromatic, electron-rich character. Graphitic nitrogen replaces a carbon atom inside the honeycomb lattice and boosts electrical conductivity. Amino nitrogen attaches to the sheet as a surface group, and oxidized nitrogen represents yet another bonding state. Each configuration changes the electronic structure of the graphene differently, and therefore each contributes differently to charge storage. The researchers set out to control that mix deliberately rather than accept whatever a decomposition reaction happened to produce.

To do so, they prepared a matrix of samples. For each of the five inorganic precursors, three graphene oxide to precursor weight ratios were tested: 1:4, 1:6, and 1:8, with the graphene oxide amount fixed at 250 milligrams. Aqueous ammonia was metered by volume instead. The resulting fifteen materials were characterized by X-ray diffraction, Raman spectroscopy, and X-ray photoelectron spectroscopy, which together revealed how crystallinity, defect density, total nitrogen content, and nitrogen bonding states shifted with both the precursor identity and the dosing level. X-ray diffraction showed a sharp peak near 26.2 degrees, corresponding to the d002 crystalline plane of restored graphitic stacking, in all samples, but the 1:6 ratio samples were consistently the most crystalline. Too little nitrogen left the sheets incompletely exfoliated and disordered; too much nitrogen over-doped them, raising the defect density until the carbon framework itself began to break down into amorphous material.

The electrochemical testing followed a careful, deliberately conservative strategy. The team first screened all samples in a three-electrode configuration on a glassy carbon electrode in 0.5 molar sulfuric acid, reporting areal capacitance rather than gravimetric values precisely because the microscopic mass loading on such a small electrode is practically impossible to measure accurately and can badly inflate specific capacitance figures. Among the 1:6 samples, areal capacitances at 5 millivolts per second ranged from 0.148 farads per square centimeter for the hydroxylamine-derived material up to 0.401 farads per square centimeter for the aqueous ammonia sample, with ammonium acetate close behind at 0.359. Cyclic voltammograms showed near-rectangular shapes, the signature of ideal electric double-layer behavior, and impedance spectroscopy confirmed that the ammonia-derived 1:6 sample had the highest conductivity and the most vertical low-frequency response, indicating the cleanest capacitive behavior.

Why did aqueous ammonia win? X-ray photoelectron spectroscopy provided the answer. The optimized sample combined a moderate total nitrogen content of 3.41 percent with the highest combined proportion of amino, pyrrolic, and graphitic nitrogen among all the materials. Nitrogen’s high electronegativity generates dipoles across the graphene plane, which improves the material’s affinity for solvated ions in the electrolyte, while graphitic nitrogen restores the conductivity that was lost when graphene oxide was first made. Raman analysis reinforced the picture: the ammonia-derived 1:6 sample showed the best ratio of graphitic to disordered carbon, the highest predicted charge carrier concentration, and the largest BET surface area at 341 square meters per gram, compared with just 91 square meters per gram for the under-doped ammonium chloride analogue, whose layered structure collapsed with insufficient nitrogen doping.

With the best material identified, the team built a symmetric two-electrode supercapacitor in a Swagelok cell. The device delivered a specific capacitance of 532 farads per gram at 0.5 amperes per gram, along with the 41 watt-hours per kilogram energy density and 375 watts per kilogram power density noted earlier. Subjected to a punishing 10,000 charge-discharge cycles at a demanding 10 amperes per gram, it retained 77.4 percent of its capacitance with a Coulombic efficiency of 97.7 percent. As a practical demonstration, the cells were wired in series to light a red LED at a combined potential of 1.96 volts. A Ragone plot traced the familiar supercapacitor trade-off, with power density climbing and energy density falling as current density rose, but the overall envelope sits comfortably in the region where these devices could complement rather than merely assist batteries.

The deepest part of the study explains the performance at the level of quantum mechanics. Using density functional theory, the team modeled graphene oxide sheets carrying each individual nitrogen configuration and computed their quantum capacitance, the intrinsic limit on charge storage imposed by the density of available electronic states near the Fermi level. Pristine graphene, with its vanishing density of states at the Fermi level, offers only about 18.5 microfarads per square centimeter. Doping changed everything. Graphitic nitrogen raised the peak to 55.6 microfarads per square centimeter, pyridinic nitrogen to 75, oxidized nitrogen to 80.5, and pyrrolic nitrogen to 91.2. But amino nitrogen was the clear standout, reaching 187.7 microfarads per square centimeter near zero volts, thanks to a large peak in the electronic density of states right at the Fermi level. All the doped structures turned out to be metallic rather than semiconducting, which is exactly what an electrode wants to be.

The theoretical ranking, amino nitrogen first and pyrrolic nitrogen second, maps directly onto the experimental results. The two best-performing precursors, aqueous ammonia and ammonium acetate, produced samples with roughly 67 percent of their nitrogen in the combined amino and pyrrolic configurations, while the worst performer, hydroxylamine, had the lowest graphitic nitrogen share and consequently the lowest conductivity and capacitance. This structure-property agreement across experiment and simulation is the study’s central contribution: it transforms nitrogen doping from a matter of trial and error into a design principle. Choose a precursor chemistry and dosing ratio that maximizes amino and pyrrolic incorporation without wrecking the carbon lattice, and the capacitance follows.

The wider implications are considerable. The synthesis uses water as the solvent, inexpensive and abundant inorganic precursors, and standard hydrothermal equipment, so the authors argue it can scale to industrial production. With graphene demand rising across energy storage, sensing, and electronics, a route that trades costly organic chemistry for ammonia and hot water could reshape the economics of doped carbon electrodes. For supercapacitors specifically, the combination of 532 farads per gram, a 1.96-volt operating window, and durable cycling over ten thousand cycles approaches the performance zone where these devices could buffer renewable grids, recover braking energy in vehicles, and power the fast-charging gadgets of the future. The recipe, in the end, is refreshingly humble: a salt, a sheet, and a sealed vessel of water, tuned one nitrogen atom at a time.

Subject of Research: Nitrogen-doped reduced graphene oxide synthesis and its quantum capacitance for high-performance supercapacitors

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

Article References: Yadav, A., Kumar, R., SanthiBhushan, B., Chandra Bhamu, K., Kang, S. G., Yan, F., & Sahoo, B. (2026). Modulating the supercapacitance of N-doped reduced graphene oxide synthesized using water-soluble inorganic precursors. Frontiers in Carbon, 5, Article 1827637. https://doi.org/10.3389/frcrb.2026.1827637

Image Credits: AI Generated

DOI: 10.3389/frcrb.2026.1827637

Keywords: supercapacitors, nitrogen-doped graphene, reduced graphene oxide, hydrothermal synthesis, quantum capacitance, energy storage, water-soluble precursors, density functional theory, graphene oxide, electrode materials, energy density, specific capacitance