Iron-impregnated cellulosic carbon as an effective electrocatalyst for seawater oxidation

Abstract

The quest for cost effective but active electrocatalysts for water oxidation is at the forefront of research towards hydrogen economy. In this regard, bamboo as biomass derived N-doped cellulosic carbon has shown potential electrocatalytic performance towards water oxidation. The impregnation of optimum metallic Fe boosts the performance further, achieving an overpotential value of 238 mV at a benchmark current density of 10 mA cm⁻². Owing to its promising OER performances in alkaline freshwater, the electrocatalyst was further explored in alkaline saline water and alkaline real seawater, exhibiting overpotentials of 272 mV and 280 mV, respectively, to reach 10 mA cm⁻² current density. Most importantly, the protective graphitic multilayer surrounding the metallic Fe allowed the electrocatalyst to demonstrate excellent durability over 30 h even at a high current density in alkaline real seawater electrolyte.

---

1. Introduction

Electrolytic hydrogen (H₂) production has become one of the most demanded eco-friendly clean energy-conversion technologies to mitigate the global energy crisis owing to its zero-carbon emission feature and high gravimetric energy density (142 kJ kg⁻¹).¹ The conventional electrochemical water electrolysis process is composed of two half-cell reactions consisting of the anodic oxygen evolution reaction (OER) and cathodic hydrogen evolution reaction (HER).² However, water electrolysis is a thermodynamically uphill chemical activation process, mainly originating from the kinetically sluggish anodic OER involving several rate-determining intermediate steps, resulting in a limited overall efficiency for clean hydrogen production. Moreover, electrochemical water splitting consumes a large amount of water for large-scale hydrogen production, which is a concern considering the freshwater scarcity on earth. In this regard, direct adoption of seawater instead of fresh water is considered a highly desirable approach for hydrogen production as 97% of the water on the earth originates from seas and oceans.³ However, the direct seawater electrolysis process is very challenging owing to the interference of multiple elements present in seawater, especially chloride ions, which cause severe electrode degradation through the chlorine evolution reaction (CER) mechanism.^4,5 Consequently, to avoid such chloride corrosion, a highly active electrocatalyst needs to be designed allowing the OER over the CER by keeping the overpotential value less than 480 mV in highly alkaline media.⁶

Furthermore, conventional OER-active electrocatalysts, such as, RuO₂ and IrO₂, are often considered expensive, making the hydrogen production costlier. In the search for cost effective but active electrocatalysts, biomass-derived carbonaceous materials as electrocatalysts have attracted much attention owing to their unique porous structure, large surface area, better electrical conductivity and presence of abundant π electrons, which could be beneficial to improve electrocatalytic activity.^7–9 However, chemical inertness of conjugated electrons of carbon nanomaterials limits their catalytic performance towards water oxidation.^10,11 Therefore, creating exposed active sites and improving the reactivity of catalytic active sites are considered effective approaches to enhance electrocatalytic activity. To date, various activation strategies, such as heterostructure engineering and doping, have been key focuses to increase the number of exposed active sites and their intrinsic reactivity for better electrochemical performances.^12–14 The incorporation of a heteroatom (N, B and P) into porous carbon alters the electronic structure for faster electron-transfer properties, better chemical stability and availability of abundant active sites, which are much influential to improve the OER activity.^15–18 For example, Gou and co-workers fabricated N and P co-doped graphene/carbon nanosheet materials (N,P-GCNS) via a pyrolysis method, which exhibited good electrocatalytic performance towards both the oxygen evolution reaction and oxygen reduction reaction.¹⁷ In another work, Yoo's group synthesized an N-doped porous carbon (N-PC) sheet-like structure from golden shower pods as a biomass resource through a solvent-free synthesis procedure.¹⁸ The optimized N-PC electrocatalyst exhibited overpotential values of 314 mV and 179 mV at the benchmarking current density of 10 mA cm⁻² for the OER and HER processes. Furthermore, the insertion of a 3d transition guest metal as a dopant can create additional multiple active sites by altering the electronic structure and coordination environment of the carbon material, thereby enhancing the intrinsic catalytic performance. Among various transition metals, iron (Fe) is recognized as an effective foreign metal that can generate an active metal centre during the water oxidation process.^19,20 For instance, Mu et al. developed N,P dual-doped carbon nanofiber-encapsulated Fe₃O₄ and FeP microporous nanoparticles through an electrospinning, carbonization, and phosphating procedure, which could act as a tri-functional electrocatalyst toward the OER, HER, and ORR processes.¹⁹ The studies revealed that while the Fe₃O₄ and N-doped carbon nanofibers acted as the main electroactive species towards both the OER and ORR reaction processes, FeP significantly enhanced the electrocatalytic HER activity in both alkaline and acidic electrolyte solutions. In fact, the carbon matrix can act as an ideal platform for stabilizing active metallic sites for catalytic applications, preventing demetallation.^21–23

Herein, we introduced metallic iron-impregnated biomass-derived cellulosic carbon as an efficient electrocatalyst for the OER by a facile two-step process involving a hydrothermal process followed by a chemical activation and pyrolysis process. The optimal electrocatalyst (BF-1) exhibited outstanding electrocatalytic activity, with an overpotential value of 238 mV to reach the benchmarking current density of 10 mA cm⁻² with a low Tafel slope value of 40 mV dec⁻¹ in alkaline fresh water electrolyte. The presence of iron in the cellulosic carbon added extra active sites and promoted the electrocatalytic OER activity by many fold. After demonstrating such excellent catalytic activity in fresh water, the catalyst was further explored in simulated as well as real seawater, and demonstrated great activity together with a promising durability for over 30 h at a very high current density.

2. Experimental section

2.1. Material synthesis

All the chemicals were analytical grade and were used directly without any further purification. In a typical synthesis process, 4.0 g of small bamboo pieces was added into a homogeneous solution of deionized water and ammonia in an 8 : 1 ratio for a total volume of 36 ml under continuous vigorous stirring at room temperature. The solution was then transferred into a Teflon-lined autoclave and kept in a hot air oven at 170 °C for 16 h. After completion of the hydrothermal reaction, the reactor was allowed to cool down to room temperature naturally. The resulting bamboo char was dried overnight at 70 °C in an oven. Afterwards, the obtained bamboo char was homogeneously mixed with KOH in a 1 : 3 ratio and placed in a tube furnace made of quartz, and the pyrolysis process was carried out at 700 °C at a ramp rate of 3° min⁻¹ and then held at 700 °C for 2 h under a continuous N₂ gas flow. Finally, the obtained product, named as BA, was washed several times with deionized water until the pH of the solution became neutral, and it was then dried overnight at 60 °C under vacuum conditions.

The incorporation of iron (Fe) into the cellulosic carbon was carried out by a similar reaction procedure, but adding iron(III) chloride anhydrous with different amounts (0.25, 0.5, and 1 g) into the homogeneous solution of DI water and ammonia during the hydrothermal treatment. Afterwards, the obtained Fe-containing bamboo char was chemically activated and pyrolyzed under the same condition as discussed for the BA product. The different content of iron-impregnated samples were abbreviated as BF (0.25 g), BF-1 (0.5 g), and BF-2 (1 g).

2.2. Structural characterization

Powder X-ray diffraction (XRD) patterns of all the samples were obtained on a Rigaku Miniflex-600 instrument using a Cu-kα source of wavelength of 1.5418 Å in the diffraction angle range of 20°–60°. The morphological structures of the as-prepared samples were investigated by field-emission scanning electron microscopy (FE-SEM, Carl Zeiss, Germany) operated at 5 kV, while the corresponding elemental composition was analysed by energy-dispersive X-ray (EDX) spectrometry associated with FESEM apparatus. Furthermore, high-resolution transmission electron microscopy (HRTEM) was performed to investigate the microstructural parameters of the samples using a FEI TECNAI 20 G2 instrument (Netherlands) operating at 200 kV. The TEM specimens were prepared by drop-casting homogeneous ink of the powder sample onto a carbon-coated Cu grid and drying overnight in a desiccator. The elemental composition and oxidation states of the samples were investigated by X-ray photoelectron spectroscopy (XPS, SCIENTA, R-3000) using monochromatic Al Kα source in a vacuum chamber (10⁻¹⁰ Torr) and the spectra analysed after Gaussian fitting. Fourier-transform infrared spectroscopy (FTIR) measurements were performed using an IR Prestige-21 instrument (SHIMADZU) in the ATR range from 400 cm⁻¹ to 4000 cm⁻¹. Surface-sensitive Raman analysis of the resulting materials was performed using an Alpha300 RAS (WITec) Raman spectrometer. Brunauer–Emmett–Teller (BET) measurements were performed by N₂ adsorption/desorption isotherms to evaluate the surface characteristics of the samples using a surface area analyzer (Quantachrome Autosorb IQ3).

2.3. Electrochemical measurements

The electrochemical OER performances of the developed electrocatalysts were evaluated at room temperature using a Metrohm Autolab204 potentiostat/galvanostat in a three-electrode set-up, consisting of Pt wire as a counter electrode, Hg/HgO as the reference electrode, and the developed electrocatalyst on Ni foam (NF) as the working electrode. The working electrode was prepared by making a homogeneous ink of the developed sample by mixing 2 mg of the ink in a 900 μl solution of isopropanol and DI water containing 100 μl solutions of PVDF under ultrasonication, and then drop-casted on the NF with a geometric area of 2 × 0.5 cm². Prior to the drop-casting procedure, the bare NF was treated by ultrasonication in 3 M HCl solution to remove the native surface oxide layer and washed several times with deionized water and ethanol. For comparison, commercially available 20 wt% RuO₂ (Sigma-Aldrich) was used to make a homogenous ink in a similar way, which was then drop-casted on NF. The optimized mass loading of all samples on the treated NF was maintained at ∼1 mg cm⁻². The electrocatalytic activity of all the developed samples was evaluated in 1 M KOH, 1 M KOH + 1 M NaCl, and 1 M KOH + seawater electrolyte solutions. In the electrochemical study, the applied potential against saturated Hg/HgO electrode was converted into the reversible hydrogen electrode (RHE) using eqn (1):

E_RHE = [E_appl + E_Hg/HgO + (0.059 × pH)]V (1)

Linear sweep voltammetry (LSV) of the developed catalysts was performed at a scan rate of 2 mV s⁻¹ in the potential range of 0–0.8 V against the Hg/HgO reference electrode for the OER process. The LSV curves of the catalyst were subjected to 80% iR correction to avoid the uncompensated series resistance of the electrochemical circuit. The electrochemical impedance spectroscopy (EIS) analysis was conducted in the frequency range of 100 kHz to 10 mHz at around 310 mV overpotential to gain insight information about the equivalent electrochemical circuit. The iR-corrected polarization curves were calculated using eqn (2):

E_iR = [E_RHE − (i × R_s)]V (2)

where R_s is the solution resistance obtained from the fitted Nyquist plot. The overpotential (η) value as well as the intrinsic catalytic kinetic parameter Tafel slope (b) of the electrodes were determined using eqn (3) and (4):

Overpotential (η) = (E_iR − 1.23)V (3)

η = a + b log j (4)

The electrochemical active surface area (ECSA) is a crucial parameter to obtain information about the real active sites of an electrocatalyst, which can be determined from the double-layer capacitance (C_dl) value assessed in the non-faradaic region in cyclic voltammetry (CV) analysis at various sweep rates following eqn (5):

where C_s is the specific capacitance of the electrode surface. The turn over frequency (TOF) is another useful parameter, and can be determined using eqn (6):^24,25

Herein, the integrated peak can be determined from the oxidation peak area of the cyclic voltammogram (CV) plot through electrochemical analysis. Chronoamperometry (CA) tests at a fixed potential were next performed to evaluate the real-time durability of the developed electrodes in highly alkaline fresh and real seawater electrolyte solutions.

3. Results and discussion

Metallic iron-impregnated cellulosic carbon was derived from bamboo as a biomass resource through a facile two-step process, comprising a hydrothermal process followed by a chemical activation and pyrolysis process, as illustrated in Fig. 1(a). Powder X-ray diffraction analysis of the developed samples was carried out in order to confirm the phase purity and information about the crystal structures of the developed samples (Fig. 1(b)). Characteristic diffraction peaks of the BA sample were observed located at about 26.0°, 43.6°, 52.94°, and 57.67°, corresponding to the (111), (100), (222), and (211) planes of graphitic-activated carbon (PDF 00-041-1487), while the remaining diffraction peaks located at 23° and 35.3° corresponded to nanocellulose (PDF 00-056-1719).^26–29 The iron impregnation in the BF samples led to the appearance of an extra peak at 45°, which could be indexed to the (111) plane for metallic Fe (JCPDS no. 06-0696). Therefore, the XRD patterns confirmed the insertion of metallic Fe impregnated in cellulosic carbon. A slight shifting of the graphitic carbon peaks towards lower angles in the BF samples compared to the BA sample also indicated a possible expansion of the interplanar spacing of graphitic carbon (Fig. S1†). The morphology of the as-prepared samples was investigated by field-emission scanning electron microscopy (FESEM) analysis. The SEM images of all the obtained samples showed a 3D network connecting randomly oriented hierarchical nanosheets in the structure together with a wrinkled texture, as represented in Fig. S2.† The EDX analysis of both the BA and BF-1 samples revealed the atomic percentages of C, O, and Fe, respectively (Table S1†). Further details on the morphological structure and microstructural parameters were obtained by transmission electron microscopy (TEM) analysis of the developed BF-1 sample. The TEM analysis confirmed the nanosheet-like morphology impregnated with iron nanoparticles with particle sizes ranging from 15–40 nm (marked with a yellow circle) as represented in Fig. 1(c) and Fig. S3.† As a consequence, the HRTEM image further showed the present of metallic iron particles (marked with a yellow dotted line) with a d-spacing value of 0.2 nm corresponding to the (110) plane of Fe in Fig. 1(d). Similarly, TEM and HRTEM analyses of the BF-2 sample also confirmed the nanosheet-like morphology and presence of metallic Fe with the (110) plane (Fig. S4†). A careful and close look at the periphery of the iron particles indicated the presence of multilayers of characteristic graphitic carbon (marked by a white dotted line). The selected area electron diffraction (SAED) pattern revealed the poor crystalline nature of BF-1 (Fig. 1(e)). Furthermore, the STEM and corresponding elemental mapping of BF-1 showed a uniform distribution of C, Fe, and O throughout the sample (Fig. 1(f)). The presence of O is believed to be related to surface oxygen due to environmental influence, which cannot be ignored.

Fig. 1 (a) Schematic representation of the synthesis procedure for the Fe-impregnated cellulosic carbon. (b) XRD patterns of all the developed samples. (c) TEM and (d) HRTEM images and (e) corresponding SAED pattern of the optimum BF-1 sample. (f) STEM image and elemental mapping of the BF-1 sample.

The characteristic nature of the developed carbon material was further verified by Raman spectroscopy analysis as represented in Fig. 2(a). In the case of the BA sample, the vibrational band in the spectrum at about 1351 cm⁻¹ could be ascribed to the presence of the D band, which originated from the disordered carbon structure, such as bond length/angle disorder and hybridization defects generated by the doping strategy.³⁰ On the other hand, the strong signal appearing at around 1584 cm⁻¹ was assigned to the graphitic (G) line, which was the E_2g vibrational mode of hexagonal graphite and was also related with the sp² hybridized of carbons in the graphite layer. These results confirmed that the biomass precursors from bamboo were successfully converted to graphitic-activated carbon, in agreement with the XRD and TEM studies. More importantly, it could be observed that the intensity of the I_D/I_G ratios of the BF samples were increased significantly compared to the BA sample, suggesting an enhancement in the structural defects after iron impregnation into the graphitic carbon.³¹ Next, nitrogen adsorption–desorption isotherms for the Brunauer–Emmett–Teller (BET) measurements were carried out to investigate the surface area of the developed graphitic carbon-based materials (Fig. 2(b)). The analysis revealed a quite large surface area value of 227 m² g⁻¹ for BA, which was increased to 284.16 m² g⁻¹ for BF-1 and 323.45 m² g⁻¹ for BF-2. Such large-surface-area graphitic-activated carbon-based materials can play an important role as electrocatalysts due to availability of large number of catalytic active sites and adsorption abilities for reactants. Fourier-transform infrared (FTIR) spectroscopy analysis of all the obtained samples was next performed to understand the functional groups and alteration of the chemical bonding, as represented in Fig. 2(c). In the case of the BA sample, there was a characteristic strong signal located at about 1540 cm⁻¹ attributed to the presence of skeletal vibration of the C=C group of the graphitic planes.³² The most intense vibrational absorption peaks located around the wavenumber of 1408 cm⁻¹ were assigned to the stretching vibration of C–C bonds.³³ The vibrational peak at the wavenumber of about 1004 cm⁻¹ was related to the stretching vibration of an alkoxy C–O group. Interestingly, a new characteristic signal appeared at 660 cm⁻¹ for the BF samples which could be attributed to Fe–O bending vibrations, further confirming the successful incorporation of iron (Fe) into the graphitic carbon layer.³⁴

Fig. 2 (a) Raman spectra of the developed samples. (b) N₂ adsorption/desorption isotherms plots for BET analysis of the samples. (c) FTIR measurements of all the samples. (d–g) High-resolution XPS spectra of (d) C 1s, (e) N 1s, (f) O 1s, and (g) Fe 2p for BA and BF-1.

The X-ray photoelectron spectroscopy (XPS) analysis of both the BA and BF-1 samples was carried out to analyze the alteration of the chemical valence states and local coordination environment after Fe insertion into the cellulosic carbon. The high-resolution C 1s spectrum of BA could be deconvoluted into two peaks at binding energies of 284.64 and 286.0 eV, attributed to the presence of carbon (C–C) with sp² hybridization and C–O bonds, respectively (Fig. 2(d)). Similar to BA, the core level C 1s spectrum of BF-1 revealed the coexistence of both C–C and C–O without any obvious changes after Fe insertion. The high-resolution N 1s spectra of both BA and BF-1 confirmed the presence of three signals with binding energies located at 398.4, 400.13, and 401.57 eV, corresponding to the presence of pyridinic-N, pyrrolic-N, and graphitic-N, respectively (Fig. 2(e)).³¹ Importantly, the presence of N caused polarization due to its higher electronegativity compared to carbon, which significantly influenced the active sites of the graphitic carbon framework and could synergistically enhance the activities for the oxygen evolution reaction process.¹⁶ In the case of both the BA and BF-1 samples, the O 1s spectra could be deconvoluted into two peaks located at about 532.16 and 533.1 eV, assigned to double-bonded oxygen to carbon (C=O) and single-bonded oxygen to carbon (C–O), respectively (Fig. 2(f)).³² The presence of Fe in BF-1 was examined considering the high-resolution Fe 2p spectrum of BF-1, as represented in Fig. 2(g).³⁵ In line with the XRD results, the carefully deconvoluted Fe 2p_3/2 peak revealed the presence of peaks at 708.2, 711.2, and 713.3 eV, corresponding to the presence of Fe⁰, Fe²⁺, and Fe³⁺ species, respectively, along with a saturated peak at 717 eV. The presence of oxidized iron species apart from Fe⁰ was believed to be due to possible surface oxidation as iron is very prone to oxidation.

3.1. Electrochemical OER performance in alkaline media

The electrochemical OER performances of the obtained electrocatalysts were initially evaluated in three-electrode configurations in alkaline 1 M KOH containing demineralized water as the electrolyte at room temperature. The electrocatalytic OER activities were compared with the catalytic performance of conventional RuO₂ drop-casted on treated NF as well as bare NF under the same electrochemical set-up. The linear sweep voltammetry (LSV) polarization curves of all the developed electrocatalysts were evaluated with 80% iR drop correction in order to avoid the uncompensated series resistance of the electrochemical circuit, as displayed in Fig. 3(a). The N-doped cellulosic carbon derived from bamboo showed better activity compared to conventional RuO₂, while a further significant enhancement of the electrocatalytic OER activities was observed after iron incorporation into the cellulosic carbon. To be precise, optimal Fe impregnation, as in BF-1, exhibited an outstanding electrocatalytic OER activity with an ultralow overpotential value of 238 mV to reach the benchmarking current density of 10 mA cm⁻². The activity trend was found to be BF-1 (238 mV) > BF (244 mV) > BF-2 (260 mV) > BA (288 mV) > RuO₂ (305 mV) > bare Ni foam (345 mV). It could thus be observed that the electrocatalytic activity was boosted with a certain Fe content in BF-1 and then decreased in BF-2 (Fig. S5†). In line with this activity trend, BF-1 also demonstrated a considerably lower overpotential value of 268 mV to reach a higher current density of 100 mA cm⁻² compared to the others (Fig. 3(b)). Moreover, the intrinsic OER kinetics of all the developed electrocatalysts were further evaluated by measuring their Tafel slope values from the LSV plots, as represented in Fig. 3(c). BF-1 exhibited a smaller Tafel slope value of 40 mV dec⁻¹ compared to BF-2 (54 mV dec⁻¹), BA (68 mV dec⁻¹), RuO₂ (90 mV dec⁻¹), and bare Ni foam (103 mV dec⁻¹), suggesting its faster reaction kinetics towards the OER process. Electrochemical impedance spectroscopy (EIS) analysis of all the developed catalysts was performed to evaluate the electron-transfer ability of the electrode during the electrochemical OER process in alkaline electrolyte solution. The fitted semi-circular Nyquist plot of the electrochemical circuit provides information about the charge-transfer resistance (R_ct) of a catalyst at the interface of the electrode and electrolyte, as displayed in Fig. 3(d). The optimized BF-1 exhibited the smallest charge-transfer resistance value of 0.89 Ω compared to the others, attributed to the improved electrical conductivity and faster charge-transfer properties of cellulosic carbon (Table S2†). The mass activity evaluation of all the developed catalysts further demonstrated BF-1 was the most effective electrocatalyst among the others (Fig. 3(e)).

Fig. 3 (a) LSV plots of all the developed samples and comparison with conventional RuO₂ and bare Ni foam in 1 M KOH electrolyte and (b) comparative bar diagram of overpotential values determined at different current densities. (c) Corresponding Tafel plots of all samples. (d) Electrochemical impedance spectroscopy (EIS) analysis of all the samples at 300 mV applied overpotential in the frequency range of 100 kHz to 10 mHz. (e) Mass activity plots of all the electrocatalysts in 1 M KOH electrolyte. (f) Chronoamperometry test of the optimized BF-1 electrocatalyst at an applied potential of 1.58 V over 24 h in 1 M KOH electrolyte.

3.2 Intrinsic catalytic activity and durability

The electrochemical active surface areas (ECSAs) of the electrocatalysts were determined from the double-layer capacitance (C_dl) value in order to gain insights into the intrinsic catalytic activity of all the developed samples. The C_dl value of all the samples was measured from the cyclic voltammetry (CV) plots in the non-faradaic region, as displayed in Fig. S6.† It was observed that BF-1 exhibited the highest C_dl value of 3.0 mF cm⁻² following the trend of BF-1 (3.0 mF cm⁻²) > BF-2 (2.75 mF cm⁻²) > BA (2.65 mF cm⁻²) > RuO₂ (2.5 mF cm⁻²), indicating its higher availability of abundant catalytic active sites compared to the others towards the OER mechanism (Fig. S7(a)†). Generally, the electrochemical active surface area is directly proportional to the double-layer capacitance value of the electrocatalyst. Agreeing with this fact, the optimized BF-1 demonstrated the highest ECSA value compared to the other developed catalysts (Fig. S7(b)†). Next, the TOF values of the catalyst were calculated from the CV curves to quantify the efficiency of the catalytic active sites (Fig. S8†). The highest TOF value of 0.08 s⁻¹ for BF-1 further indicated its best electrocatalytic activity due to the large number of active sites in BF-1. Besides the electrocatalytic activity, the electrochemical durability is another important factor to evaluate the practical ability of the developed electrocatalyst in an alkaline electrolyte medium. Thereby, a chronoamperometry (CA) study of optimized BF-1 electrocatalyst was carried out to verify its stability during continuous OER performance. The CA study of the BF-1 electrocatalyst demonstrated its excellent stability over 24 h at 1.58 V vs. RHE applied potential, maintaining a high current density of 100 mA cm⁻² without significant current degradation (Fig. 3(f)). The LSV plot of the BF-1 catalyst before and after the stability test further revealed there was only a small changes in activity even after 24 h continuous OER operation in alkaline solution (Fig. S9(a)†).

Certainly, the developed pristine cellulosic carbon without and with Fe impregnation showcased better electrocatalytic activity compared to reported Fe-based and carbon-based electrocatalysts, as compared in Table S4.†

3.3. OER performance in alkaline saline and real seawater

Having demonstrated outstanding electrocatalytic activity in alkaline freshwater, the electrochemical OER performance of the BF-1 was further examined in alkaline saline and real seawater electrolytes. LSV polarization plots of BF-1 were performed in different electrolyte media, such as 1 M KOH + 1 M NaCl and 1 M KOH + seawater electrolyte and compared with 1M KOH containing freshwater, as displayed in Fig. 4(a). Due to the interference of various elements available in seawater, BF-1 exhibited a slightly higher overpotential value of 280 mV in alkaline seawater as anticipated compared to alkaline fresh water to reach a current density of 10 mA cm⁻² (Table S3†). However, importantly, the overpotential value of BF-1 remained low enough (336 mV) within the limiting overpotential value of 480 mV to reach the higher current density of 100 mA cm⁻² in alkaline real seawater, suggesting its favourable anodic OER electrocatalytic activity over the CER (Fig. 4(b)). The EIS study of BF-1 showed a lower charge-transfer resistance (R_ct) value in both alkaline fresh water as well as saline water, as represented in Fig. 4(c). The Tafel slope values of BF-1 were also found to be small enough in alkaline saline water and alkaline real seawater, indicating faster reaction kinetics in both electrolytes (Fig. 4(d)). Most importantly, the CA study of BF-1 demonstrated an outstanding durability over 30 h at an applied potential of 1.65 V vs. RHE, keeping a current density of 80 mA cm⁻² without much alteration in the alkaline seawater electrolyte (Fig. 4(e)). The LSV polarization plots before and after the CA test showed comparable electrochemical performance without any significant degradation, verifying the great stability of the electrocatalyst in such a chloride-containing harsh medium (Fig. S9(b)†).

Fig. 4 (a) LSV plots of the optimized BF-1 electrocatalyst in different electrolyte media and (b) comparative bar diagram of the overpotential values obtained at different current densities. (c) EIS spectra of BF-1 in different alkaline electrolytes at 300 mV overpotential in the frequency range of 100 kHz to 10 mHz. (d) Tafel plots of BF-1 in different electrolyte media. (e) CA test of BF-1 at a high applied potential of 1.65 V vs. RHE over 30 h towards the OER mechanism in 1 M KOH + seawater electrolyte.

Careful evaluation of the stability of BF-1 after a longer OER stability test in alkaline seawater was also carried out using XPS analysis (Fig. 5). Impressively, no noticeable changes were found, and the presence of N and Fe along with C, in BF-1, was well confirmed, after the long chronoamperometry test in highly alkaline medium. In fact, the high-resolution Fe 2p peak also showed the presence of Fe⁰ along with its other species in the form of Fe²⁺ and Fe³⁺, confirming the great stability of the developed electrocatalyst in chloride-containing harsh media. Moreover, the morphological and structural stability of the BF-1 electrode after the stability test were examined using FESEM and XRD analyses (Fig. S10†). Though the XRD peaks for electrocatalysts found to be very small compared to the strong peaks for Ni foam due to the small sample content drop-casted on the Ni foam, careful analysis indicated the stable composition of BF-1, well supported by the EDX mapping showing the presence of all the elements, same as in the XPS analysis after the stability test. It is believed that the presence of graphitic carbon surrounding the metallic iron offers much needed stability.

Fig. 5 High-resolution XPS spectra for (a) C 1s, (b) N 1s, (c) O 1s, and (d) Fe 2p for BF-1 before and after the chronoamperometry test in alkaline seawater electrolyte.

4. Conclusions

In conclusion, we successfully synthesized iron-impregnated cellulosic carbon with a graphitic nature as an efficient electrocatalyst by a two-step process involving a hydrothermal process followed by a chemical activation and pyrolysis process. While the N-doped cellulosic carbon itself showed better electrocatalytic performances towards the OER, the optimum impregnation of metallic Fe boosted the performance further. The electrocatalyst was found to have much potential for exploration in chloride-containing alkaline saline water or alkaline real seawater, exhibiting overpotentials of 272 and 280 mV, respectively, while in alkaline freshwater only a 238 mV overpotential was required to reach the benchmarking current density of 10 mA cm⁻². Impressively, the electrocatalyst also demonstrated excellent durability over 30 h even at a higher current density in alkaline real seawater electrolyte. The presence of metallic iron well protected by the graphitic carbon enabled it to achieve great activity along with longer durability in alkaline real seawater electrolyte. Having such desirable properties, the cost-effective bamboo-derived iron-impregnated cellulosic carbon is believed to have great potential for practical implementation for efficient seawater oxidation.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The work on development of carbonaceous materials is an output of project MLP 230012 and the electrocatalytic seawater splitting was carried out under HCP44/01 both funded by Council of Scientific & Industrial Research (CSIR, India). S. K. thanks University Grant Commission (UGC, India) for providing fellowship to carry out her PhD program.

References

1. S. Khatun, K. Shimizu, S. Singha, R. Saha, S. Watanabe and P. Roy, Chem. Eng. J., 2021, 426, 131884.
2. S. Pal, T. Ahmed, S. Khatun and P. Roy, ACS Appl. Energy Mater., 2023, 6, 7737–7784.
3. F. Dionigi, T. Reier, Z. Pawolek, M. Gliech and P. Strasser, ChemSusChem, 2016, 9, 962–972.
4. S. Khatun, H. Hirani and P. Roy, J. Mater. Chem. A, 2021, 9, 74–86.
5. S. Khatun, S. Pal, N. Sinha, C. Das, T. Ahmed and P. Roy, Chem. Commun., 2023, 59, 4578–4599.
6. S. Khatun, K. Shimizu, S. Pal, S. Nandi, S. Watanabe and P. Roy, Small, 2024, 2402720.
7. Q. Chen, X. Tan, Y. Liu, S. Liu, M. Li, Y. Gu, P. Zhang, S. Ye, Z. Yang and Y. Yang, J. Mater. Chem. A, 2020, 8, 5773–5811.
8. G. Zhang, X. Liu, L. Wang and H. Fu, J. Mater. Chem. A, 2022, 10, 9277–9307.
9. Y. Feng, J. Jiang, Y. Xu, S. Wang, W. An, Q. Chai, U. H. Prova, C. Wang and G. Huang, Carbon, 2023, 211, 118105.
10. Y. Xu, M. Kraft and R. Xu, Chem. Soc. Rev., 2016, 45, 3039.
11. W. Li, C. Wang and X. Lu, J. Mater. Chem. A, 2021, 9, 3786.
12. J. Chen, Q. Long, K. Xiao, T. Ouyang, N. Li, S. Ye and Z. Q. Liu, Sci. Bull., 2021, 66, 1063–1072.
13. Z. Wang, J. Huang, L. Wang, Y. Liu, W. Liu, S. Zhao and Z. Q. Liu, Angew. Chem., Int. Ed., 2022, 61, e202114696.
14. W. K. Han, J. X. Wei, K. Xiao, T. Ouyang, X. Peng, S. Zhao and Z. Q. Liu, Angew. Chem., Int. Ed., 2022, 61, e202206050.
15. H. Wang, T. Maiyalagan and X. Wang, ACS Catal., 2012, 2, 781–794.
16. J. Kim, J. Park, J. Lee, W. G. Lim, C. Jo and J. Lee, Adv. Funct. Mater., 2021, 31, 1–9.
17. R. Li, Z. Wei and X. Gou, ACS Catal., 2015, 5, 4133–4142.
18. C. Sathiskumar, S. Ramakrishnan, M. Vinothkannan, A. R. Kim, S. Karthikeyan and D. J. Yoo, Nanomaterials, 2020, 10, 76.
19. M. Wang, C. Zhang, T. Meng, Z. Pu, H. Jin, D. He, J. Zhang and S. Mu, J. Power Sources, 2019, 413, 367–375.
20. P. Li, H. Wang, W. Fan, M. Huang, J. Shi, Z. Shi and S. Liu, Chem. Eng. J., 2021, 421, 129704.
21. Q. Ma, H. Jin, J. Zhu, Z. Li, H. Xu, B. Liu, Z. Zhang, J. Ma and S. Mu, Adv. Sci., 2021, 8, 2102209.
22. Y. Zhou, R. Lu, X. Tao, Z. Qiu, G. Chen, J. Yang, Y. Zhao, X. Feng and K. Mullen, J. Am. Chem. Soc., 2023, 145, 3647–3655.
23. K. Y. Xiong, L. W. Shen, Y. Wang, Y. Liu, M. X. Hu, J. Ying, Y. X. Xiao, L. Shen, G. Tian and X. Y. Yang, Chem. Commun., 2024, 60, 7499–7502.
24. M. S. Burke, M. G. Kast, L. Trotochaud, A. M. Smith and S. W. Boettcher, J. Am. Chem. Soc., 2015, 137, 3638–3648.
25. J. Jin, Y. Yin, Y. Hu, Y. Zheng, H. Liu, X. Wang, P. Xi and C. H. Yan, Angew. Chem., 2024, 136, e202313185.
26. N. Sofyan, S. Alfaruq, A. Zulfia and A. Subhan, Jurnal Kimia dan Kemasan, 2018, 40, 9.
27. A. Taratayko, E. Kolobova and G. Mamontov, Catalysts, 2022, 12, 1393.
28. Y. Ma, X. Chai, H. Bao, Y. Huang and W. Dong, PLoS One, 2023, 18, e0281142.
29. L. Mahardiani, R. Ilfama, S. Saputro, S. H. Pranolo and P. W. Septianing, J. Phys.: Conf. Ser., 2021, 1912, 012015.
30. A. Zhang, C. Wang, Q. Xu, H. Liu, Y. Wang and Y. Xia, RSC Adv., 2015, 5, 26017–26026.
31. Y. Zhou, Y. Sun, H. Wang, C. Zhu, J. Gao, D. Wu, H. Huang, Y. Liu and Z. Kang, Inorg. Chem. Front., 2018, 5, 2985–2991.
32. S. Drewniak, R. Muzyka, A. Stolarczyk, T. Pustelny, M. Kotyczka-Morańska and M. Setkiewicz, Sensors, 2016, 16, 103.
33. A. Sabarudin, R. Wahid, F. C. Nalle, R. A. Shobirin and D. J. D. H. Santjojo, Rasayan J. Chem., 2017, 10, 1261–1270.
34. G. Asab, E. A. Zereffa and T. Abdo Seghne, Int. J. Biomater., 2020, 2020, 4783612.
35. S. Khatun and P. Roy, Chem. Commun., 2020, 56, 7293–7296.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4dt01485e

---