Heavily Graphitic‑Nitrogen‑Doped High‑Porosity Carbon from Kidney Bean Biomass for Robust Oxygen Reduction Electrocatalysis

Abstract

Large‑scale fabrication of active, stable, porous carbon catalysts for the oxygen reduction reaction (ORR) from protein‑rich biomass has become a focal point in fuel‑cell research. We present a streamlined, two‑step pyrolysis approach that couples zinc chloride activation and acid leaching to produce nitrogen‑doped, high‑porosity nanocarbons. Kidney bean, a protein‑rich and readily available feedstock, serves as the sole carbon‑nitrogen source. The resulting material, KB350Z‑900, delivers ORR electrocatalytic performance that rivals or exceeds that of commercial 20 wt % Pt/C, while exhibiting superior durability and methanol tolerance. These findings underscore the potential of biomass‑derived, graphitic‑nitrogen‑rich carbons as cost‑effective, high‑performance ORR catalysts.

Background

Platinum (Pt)-based electrocatalysts dominate fuel‑cell technology but are hampered by high cost, limited availability, and poor methanol tolerance during the ORR [1]. Consequently, the scientific community has focused on developing highly active, durable, and inexpensive alternatives. Heteroatom‑doped porous carbons (HDPC) have emerged as promising metal‑free catalysts, offering high activity, robustness, and renewability [2‑6]. However, most HDPCs are synthesized via costly chemical routes or complex templates, hindering their large‑scale adoption [7,8]. Therefore, a rational, scalable, and high‑yield synthesis strategy remains a critical challenge for next‑generation ORR catalysts.

Protein‑rich biomass such as nori [9], sweet potato vine [10], pomelo peel [11], enoki mushroom [12], coprinus comatus [13], and Lemna minor [14] has been shown to serve as single‑source precursors for HDPCs. Recent work demonstrated that high‑temperature carbonization of fish‑scale waste in the presence of zinc chloride generates a 3‑D porous network with superior ORR activity [6]. Importantly, a preliminary pretreatment step can enhance both the carbon structure and the nitrogen incorporation efficiency. Building on these insights, we report the first synthesis of heavily graphitic‑nitrogen‑doped, high‑porosity carbons (KB350Z‑900) derived from white kidney bean (KB) biomass through a two‑step pyrolysis, zinc chloride activation, and acid‑leaching sequence. White kidney bean is abundant, inexpensive, and rich in protein (20–30 % in dry mass). To our knowledge, no ORR study of KB‑derived doped carbons has been reported. Zinc chloride promotes rapid dehydration and dehydroxylation, driving the formation of micro/mesopores and a nitrogen‑self‑doped, highly porous carbon matrix. The resulting catalyst displays exceptional electrocatalytic activity, long‑term durability, and methanol tolerance, positioning it as a viable alternative to Pt‑based systems in alkaline media.

Methods

White kidney bean was first washed with deionized water and dried at 80 °C in a vacuum oven. It was then pretreated under flowing nitrogen at 350 °C for 2 h (KB350 precursor), ensuring complete protein decomposition. Although the fastest decomposition of KB occurs near 300 °C (see Fig. S1), 350 °C was chosen to surpass the tyrosine decomposition temperature (344 °C), the highest among amino acids in the protein. The KB350 precursor was mechanically mixed with zinc chloride (ZnCl₂) at a 1:1 mass ratio via ball‑milling (500 rpm). The mixture was pyrolyzed in a tubular furnace at 700, 800, 900, or 1000 °C for 2 h with a heating rate of 10 °C min⁻¹, yielding KB350Z‑X (X = 700, 800, 900, 1000). For comparison, KB‑Z‑900 was prepared by mixing raw KB with ZnCl₂, and KB900 was produced by direct pyrolysis of KB at 900 °C. All samples were subsequently acid‑leached in 0.5 mol L⁻¹ HCl for 2 h to remove Zn residues and metallic impurities before electrochemical testing.

Raman spectra were recorded using a Renishaw inVia spectrometer (λ = 514.5 nm). FE‑SEM images were obtained on a Hitachi UHR S4800. HR‑TEM was performed on a FEI Tecnai F30 (300 kV). XPS analyses employed a Kratos XSAM800. Nitrogen adsorption/desorption was measured on a Micromeritics ASAP 2010 at 77 K.

Electrochemical measurements were conducted on a Zennium‑E workstation (Germany) with a conventional three‑electrode setup: glass‑carbon rotation disk electrode (GC‑RDE, Φ = 4 mm), saturated calomel electrode (SCE), and graphite rod (Φ = 0.5 cm). Working electrodes were prepared by depositing 5.0 µL of a 10 mg mL⁻¹ dispersion onto the GC‑RDE and drying. Mass loading was ~400 µg cm⁻². All potentials vs. SCE were converted to RHE.

Results and Discussion

Raman spectra (Fig. 1a) reveal that KB350Z‑900 possesses the lowest I_D/I_G ratio (0.85) among the samples, indicating a higher graphitic degree compared to KB900 (0.94) and KB‑Z‑900 (0.88). The presence of ZnCl₂ during activation further enhances graphitization, while the 350 °C pretreatment promotes incorporation of graphitic nitrogen, as reflected in the reduced I_D/I_G ratio.

N₂ adsorption–desorption isotherms (Fig. 1b) display a type H₂ hysteresis loop, confirming mesoporosity. KB‑Z‑900 and KB350Z‑900 exhibit BET surface areas of 380 m² g⁻¹ and 1132 m² g⁻¹, respectively, with a total pore volume of 0.62 m³ g⁻¹ for KB350Z‑900. The substantial meso‑ and macroporosity (~664 m² g⁻¹) enhances active‑site exposure and oxygen diffusion during ORR. TEM images (Fig. 1c,d) corroborate the presence of abundant micro/mesopores and defected edges associated with high nitrogen content, providing efficient ORR active sites.

XPS confirms successful nitrogen doping: surface nitrogen contents are 1.23 at.% (KB‑900), 1.92 at.% (KB‑Z‑900), and 2.70 at.% (KB350Z‑900). The N 1s spectra of KB350Z‑900 show only pyridinic‑N (398.5 eV) and graphitic‑N (401.1 eV) peaks, with graphitic‑N accounting for 88.8 at.% of total nitrogen—higher than in KB‑Z‑900 and KB‑900. The absence of oxidized‑N species indicates that ZnCl₂ activation and the two‑step pyrolysis suppress N‑oxidation.

CV curves (Fig. 3a) in N₂ vs. O₂‑saturated KOH reveal that KB350Z‑900 achieves the most positive peak potential (0.90 V) and highest peak current density, attributable to its high graphitic‑N content. LSV measurements (Fig. 3b) show that KB350Z‑900’s half‑wave potential and limiting current approach those of a commercial 20 wt % Pt/C catalyst. Tafel analysis yields a slope of 143 mV dec⁻¹ for KB350Z‑900, indicating favorable kinetics, although slightly higher than Pt/C. The ORR activity order—KB350Z‑900 > KB350Z‑800 > KB350Z‑1000 > KB350Z‑700—underscores the optimal 900 °C pyrolysis temperature for balancing porosity, nitrogen content, and graphitization.

Koutecky–Levich plots (Fig. 4c) at varying rotation rates confirm first‑order kinetics and four‑electron transfer pathways, with average electron numbers of 3.93 (KB‑Z‑900) and 3.98 (KB350Z‑900). Accelerated aging tests (5000 CV cycles, 0.2–1.2 V vs. RHE) demonstrate a mere 2 mV shift in KB350Z‑900’s half‑wave potential versus a 55 mV shift for Pt/C, highlighting superior durability. Amperometric i–t curves at 0.9 V (inset Fig. 4d) confirm negligible methanol oxidation on KB350Z‑900, affirming methanol tolerance.

Conclusions

We have developed a scalable, two‑step pyrolysis protocol that transforms protein‑rich kidney bean biomass into a high‑porosity, heavily graphitic‑nitrogen‑doped carbon (KB350Z‑900). This material boasts a BET surface area of 1132 m² g⁻¹, a pore volume of 0.62 m³ g⁻¹, and a graphitic‑N content of 88.8 at.%. Its ORR performance rivals that of commercial 20 wt % Pt/C, while offering markedly better durability and methanol tolerance in alkaline media. The synergy of ZnCl₂ activation, acid leaching, and a preliminary 350 °C pretreatment underpins the exceptional catalytic activity. These results pave the way for cost‑effective, biomass‑derived ORR catalysts in fuel‑cell applications.