The growing demand for clean and sustainable energy sources on a global scale has induced significant research efforts in the field of electrochemical energy storage. Simultaneously required are advances in energy conversion devices and the development of efficient and multipurpose energy storage solutions for the ubiquitous adoption of renewable energy systems, electric transportation, and smart grid integration. Batteries and supercapacitors play a crucial role in the efficient capture, storage, and delivery of electrical energy on demand (Yadlapalli et al., 2022). They provide an essential connection between intermittent energy sources, such as solar and wind, and the constant power demands of contemporary society.

As a result, a great deal of focus has been placed on the development of supercapacitor electrode materials to improve energy storage performance and meet the high demand for modern electronic devices. Electrode materials based on transition metal compounds, such as metal oxides, metal hydroxides, metal sulfides, and metal nitrides, are regarded as optimal for highly efficient electrochemical capacitors and hybrid supercapacitors (HSCs) (Zhi et al., 2013; Augustyn et al., 2014; Chi et al., 2018; Moyseowicz, 2019). Transition metal nitrides have the ability to substitute standard materials in a variety of applications, including electrical devices, gas sensing, ceramics, and environmental remediation. Due to their low density, large surface area, electrical properties, enhanced chemical activity, and high conductivity, nanoparticles with diverse morphologies and structures have recently attracted the interest of researchers (Zhou et al., 2019; Idrees et al., 2021). Nevertheless, similar to other conversion-type materials, obstacles such as repeated volume variation, a highly air-sensitive surface, and simple oxidation significantly degrade their properties (Adalati et al., 2022). Among the various metal nitrides, Fe₂N is a very promising electroactive material because of its high electrical conductivity, low cost, easy preparation and environmental friendliness, however, limited amount of publications on its application in energy storage devices (Xu et al., 2017; Śliwak et al., 2017; Feng et al., 2023; Zhao et al., 2023).

The choice of iron is primarily motivated by its remarkable electrochemical capabilities and its abundance in nature, making it an economically feasible choice. Iron exhibits many valence states, such as Fe⁰, Fe²⁺, Fe³⁺, etc., and possesses a diverse range of redox chemistry (Feng et al., 2023). These characteristics contribute to its ability to achieve high specific capacitance values. The majority of transition metal oxides/hydroxides are commonly utilized as positive electrodes, but iron oxides/hydroxides/nitrides exhibit a stable and extensive operating range in the negative potential. This characteristic renders them highly promising as negative electrodes for asymmetric or hybrid supecapacitors (Zhang et al., 2016).

Carbon materials are extensively employed in energy storage devices, particularly in electric double-layer capacitors (EDLCs) (Borenstein et al., 2017). Carbon-based materials are a very interesting group in materials science for their superior mechanical strength, structural stability, and electrical conductivity, while also being lightweight and cost-effective (Moyseowicz et al., 2021). Furthermore, carbon cloth, which is a crisscrossed interweaving structure of carbon threads that are composed of carbon fibers or nanofibers, is the most widely used current collector in the field of supercapacitors, particularly flexible supercapacitors (Chen et al., 2018). They increase the surface area of the electrodes and offer flexibility (Gong et al., 2021).

Current research on supercapacitors indicates that new strategies must be implemented to achieve comparable energy density to lithium-ion batteries. Consequently, a new strategic design for HSCs has been developed (Jin et al., 2008; Dubal et al., 2015; Moyseowicz and Moyseowicz, 2021). In such a device, the electrode configuration consists of one capacitive or pseudo-capacitive electrode and one battery-type electrode. By combining the benefits of supercapacitor and battery electrodes, this type of electrode configuration in supercapacitors makes it possible to get high energy and power densities (Cherusseri et al., 2019). HSCs employing battery electrodes can bridge the divide between standard supercapacitors and lithium-ion batteries.

In the present study, we investigated the influence of the synthesis procedure on the morphology, crystalline structure, and electrochemical performance of Fe₂O₃ deposited on the carbon cloth substrate and their iron nitride analogues. Two environmentally friendly, scalable, and straightforward synthesis methods, including hydrothermal treatment and direct electrodeposition, were utilized. The results show that the precursor synthesis for iron nitride has a direct effect on the iron-based nanoparticle coating properties, which greatly impacts the electrode electrochemical behavior in a 6 mol L^-1 KOH electrolyte. The binder-less electrodes prepared from the direct electrodeposition method displayed much superior Faradaic behavior, yielding a specific capacity of up to 186 mAh g^-1, greater than the composite electrode containing Fe₂N from the hydrothermal method.

In Figure 1 the comparison between the morphologies of the iron-based films deposited by the two techniques, namely from electrodeposition and hydrothermal/drop casting technique is presented, both before and after nitridation. Comparison between Figures 1A, B show the distinct difference between Fe₂O₃ obtained by the two methods. As can be seen, the morphology of the electrodeposited film is composed of continuously interconnected nanosheets, while hydrothermal technique led to formation of the film in the form of rather uniform nanoparticles. Partially uncovered carbon cloth surfaces also suggests that the deposition of Fe₂O₃ on the substrate is not homogeneous within the sample.

Figure 1C reveals that thermal treatment of the Fe₂O₃@hCC-E sample in an ammonia atmosphere resulted in a radical change in the film morphology. As can be seen, a porous interconnected structure was converted into polydisperse micro- and nanoparticles evenly distributed on the carbon cloth surface. On the other hand, the thermal conversion of Fe₂O₃@hCC-HT in ammonia resulted in the aggregation of nanoparticles and formation of bulk particles (Figure 1D). As can be seen, the nitridation procedure, even if it is performed under the identical conditions, had a very different effect on two oxide materials and the morphology of nitride samples obtained by the two techniques is also distinctly different.

Furthermore, XRD diffraction patterns (Figure 2) present very different structures for the obtained samples. As can be seen, the iron oxide sample obtained by the hydrothermal technique shows a typical pattern characteristic for hematite (α-Fe₂O₃—JCPDS 33–0664) phase with reflexes located at 2Θ of 24.2°, 33.2°, 35.7°, 40.8°, 49.5°, 54.1°, 57.5°, 62.3°, 64.0°, 71.9° and 75.4° (Geng et al., 2012; Wu et al., 2013; Moyseowicz et al., 2017). The deposited film is definitely well-crystallized due to sharp feature of the reflexes. Besides that, broad signals located at 2Θ of c.a. 26° and 44° can be seen in this diffractogram, that are characteristic for carbon cloth (Walendzik et al., 2023). The nitridation of the Fe₂O₃@hCC-HT sample also resulted in the formation of well-crystallized product. Diffractogram of Fe₂N@hCC-HT shows, besides signals derived from carbon cloth, a diffraction pattern typical for ζ-Fe₂N phase (JCPDS 50–0958), with reflexes located at 2Θ of 35.4°, 37.4°, 40.1°, 40.9°, 43.0°, 56.9°, 68.1° and 76.0° (Śliwak et al., 2017). The narrow width of the reflexes suggests the big nitride crystallite size. On the other hand, diffractograms of both samples prepared with aid of electrodeposition technique are very different from those of materials prepared hydrothermally. Despite the presence of the thick electrodeposited film in the Fe₂O₃@hCC-E sample (as evidenced by SEM), diffractogram of this sample has only signals that can be attributed to carbon cloth. This leads to the conclusion that the deposited film is amorphous or sub-nanocrystalline (Valvoda, 1996). What is more, an iron nitride (Fe₂N@hCC-E) formed upon amination of this sample shows similar diffractogram, therefore the nanoparticles that can be seen in its SEM image are also of amorphous structure.

In order to further characterize the chemical states of elements on the surfaces of the 4 electrodes, XPS technique was applied. As can be noticed in Table 1, iron is present in all samples, indicating that in all samples, iron-based films are formed. The iron concentration in the samples obtained by hydrothermal treatment is evidently higher than in electrodeposited materials, suggesting that more active material was deposited by the former technique. Fe/O atomic ratio for both samples with iron oxide films is similar (0.54 and 0.49 for hydrothermal and electrodeposition techniques, respectively), indicating similar surface chemical states of both oxides.

Similar atomic concentrations of nitrogen can be found in both samples obtained after nitridation. For both samples Fe/N ratio (of 6.1 for Fe₂N@hCC-HT and 2.8 for Fe₂N@hCC-E) is higher than stoichiometric for Fe₂N (2) suggesting that in both materials iron can partially be found in the surface in different form. It is possible that upon the thermal treatment, iron nitride with lower nitrogen content was formed (Park et al., 2016). It can be pointed out here, that metal nitrides usually undergo surface oxidation with the formation of metal oxides (Dong et al., 2017).

Furthermore, high resolution XPS spectra in the Fe 2p and N 1s regions are presented in Figure 3 and Figure 4. In all spectra distinct Fe 2p_3/2 and Fe 2p_1/2 peaks can be seen at binding energies of 711.6 and 725.4 eV, respectively, that are characteristic for α-Fe₂O₃ phase. Typical for this phase are also the satellite features located at binding energies of 719.5 and 733.0 eV (Cao et al., 2015; Yan et al., 2017). The fact that both oxide and nitride materials possess nearly identical XPS spectra in the Fe 2p region can be explained by the oxidation of surface Fe₂N with the formation of thin Fe₂O₃ layer. XPS is a surface sensitive technique which analyzes only a few atomic monolayers, so majority of the signal comes from this oxidized layer. Surface oxidation of Fe₂N into oxide phase was widely discussed in the literature (Dong et al., 2017). As oxidized layer is very thin, therefore its diffraction reflexes cannot be distinguished in XRD patterns of nitride samples.

High resolution N 1s spectra are presented in Figure 4 for both nitride samples. The spectra can be deconvoluted into two peaks. In both spectra, the most intensive component is located at the binding energy of 399.1 eV and can be attributed to metal-nitrogen bonds or pyridinic N, as their binding energy values are almost the same (Wu et al., 2011; Peng et al., 2013; Yu et al., 2018; Kohila Rani et al., 2020). The other peak can be distinguished at 404.0 and 403.4 eV for Fe₂N@hCC-HT and Fe₂N@hCC-E, respectively. Signals in this spectral region (between 403 and 406 eV) are usually attained to nitrogen-oxygen bonds (Arrigo et al., 2010; Kordek-Khalil et al., 2022). Characterizations of all samples show that the two nitride samples are both composed of Fe₂N phase covered by thin Fe₂O₃ layer generated by atmospheric oxidation of the nitrides. What differs between the samples is the film morphology as well as structure, which in the case of the sample prepared by hydrothermal technique is crystalline with distinct XRD reflexes and is amorphous for the electrodeposited film. Similar is the case with source metal oxides, which also possess similar surface characteristics but differ in morphology and structure. In general, electrodeposition leads to the formation of amorphous films, and the hydrothermal method forms well-crystallized nanocrystalline products.

Figure 5 presents HRTEM images of the samples after the nitridation procedure. As can be seen in the bright-field images, the Fe₂N@hCC-HT sample is composed of smaller nanoparticles than Fe₂N@hCC-E, which is consistent with SEM images. Furthermore, high-resolution TEM images of the samples were also acquired to get an insight into the structure of the materials. As can be seen in Figure 5B, the sample Fe₂N@hCC-E possesses small crystalline domains (with a diameter of ca. 10 nm) in a rather amorphous matrix. The interplanar distance of 0.210 nm in these domains corresponds to the (011) crystalline plane of hexagonal Fe₂N (Jiang et al., 2019), which is also confirmed by the SAED ring (Supplementary Figure S1). On the other hand, Fe₂N@hCC-HT sample is evidently more crystalline with a fringe spacing of 0.25 nm (Figure 5D) and can be attributed to Fe₂O₃, which based on XPS results, is a product of surface oxidation of Fe₂N (Chen et al., 2016). Additionally, the SAED ring patterns also implies the Fe₂O₃ presence on the surface of the Fe₂N@hCC-HT (Supplementary Figure S1).

Figure 6 displays the findings of the CV measurements, which reveal distinct differences in the electrochemical behavior between iron oxides and iron nitrides, as well as comparision between materials from electrodeposition and hydrothermal/drop casting method. The iron-based compounds exhibit clear redox peaks when a lower cutoff potential is used (i.e., below −1.0 V vs. Hg|HgO). The Fe₂O₃@hCC-HT electrode shows poor electrochemical behavior (Figure 6A), with a very weak peaks of an Fe³⁺ ⇋ Fe²⁺ redox pair and a small peak related to Fe²⁺ reduction to metallic Fe⁰ at a potential of −1.05 V vs. Hg|HgO electrode (Zhang et al., 2016). However, the Fe₂O₃ obtained from the electrodeposition process presents improved electrochemical performance compared to its hydrothermal counterpart (Figure 6B). The CV curve of Fe₂O₃@hCC-E shows the cathodic peak attributed to Fe²⁺ reduction to metallic Fe⁰, however due to its nanosheet-like morphology, the additional peak appeared at −0.71 V vs. Hg|HgO electrode, related to the conversion of the intermediate phase of adsorbed iron hydroxide to Fe²⁺ (Li et al., 2015). The effect of a high temperature nitridation of the Fe₂O₃@hCC electrodes is very pronounced and greatly improves the electrochemical response of converted iron oxides to nitrides. Due to the increase in conductivity of the Fe₂N form, the CV profile of the Fe₂N@hCC-HT presents distinct redox pair related to the transition of Fe³⁺ to Fe²⁺ at the potentials value of −0.97 V vs. Hg|HgO for cathodic and −0.64 V vs. Hg|HgO for anodic peak (Xu et al., 2017; Zhao et al., 2023), therefore indicating a battery-type behavior of the Fe₂N in an alkaline electrolyte. Furthermore, the CV curve revealed that during electrochemical measurement, the Fe₂N@hCC-E sample undergoes further reduction to metallic iron, showing a peak above 1.0 V vs. Hg|HgO, which may influence a Fe³⁺ ⇋ Fe²⁺ redox potential, shifting it towards more positive values by 70 mV (Li et al., 2015). Finally, even at a low scan rate of 5 mV s⁻, the peak correlated to the hydrogen evolution reaction is not observed for the Fe₂N@hCC samples, due to the increased hydrogen evolution potential of iron nitride (Theerthagiri et al., 2020).

The GCD measurements confirm the results obtained from the CV measurements. The Fe₂O₃@hCC electrodes show a distorted, quasi-triangular shape, with an linear part above −0.8 V vs. Hg|HgO, which is mostly related to the charge stored in the Fe²⁺ to Fe⁰ transition (Figures 6C, D). Moreover, the time required for discharge of the Fe₂O₃@hCC-E is over 6 times longer than that of Fe₂O₃@hCC-HT at a current density of 0.5 A g^-1. The nitridation of the Fe₂O₃@hCC electrodes changes significantly the GCD profiles, as visible plateaus are observed, which correspond to the redox peaks of cyclic voltammograms at −0.71 ÷ −0.64 V and −1.05 ÷ −0.97 V vs. Hg|HgO (Figures 6E, F), thus confirming the Fe₂N battery-like electrochemical behavior (Simon et al., 2014; Xu et al., 2017).

The relationships between the specific capacity and scan rate are presented in Figure 7A. The highest specific capacity at a scan rate of 1 mV s⁻¹ is achieved by the Fe₂N@hCC-E sample, with a value of 186 mAh g⁻¹, followed by Fe₂O₃@hCC-E (177 mAh g⁻¹), showing the superiority of the electrodeposition method over the hydrothermal synthesis combined with drop casting technique. The specific capacity of Fe₂N@hCC-HT was 112 mAh g⁻¹, 66% less than its counterpart from the electrodeposition procedure, while the lowest specific capacity value of 43 mAh g^-1 was measured for Fe₂O₃@hCC-HT sample. Furthermore, from the discharge time of the GCD profiles, the specific capacity values (Q_s) can be calculated, accordingly to the Equation 2 and at a current density of 0.5 A g^-1 the Q_s were 22, 126, 98 and 135 mAh g^-1, Fe₂O₃@hCC-HT, Fe₂O₃@hCC-E, Fe₂N@hCC-HT and Fe₂N@hCC-E, respectively, corresponding to the performance between 2 and 5 mV s^-1 from the CV measurements. The aggregated, spherical-like morphology does not provide enough active sites for redox reactions; thus, iron-based films from hydrothermal treatment have lower charge storage capability. Moreover, the small nanoparticle size of a thin iron nitride film in the Fe₂N@hCC-E sample and the amorphous nature of the iron oxide precursor led to the improved electrochemical response (Zhang et al., 2016; Eftekhari and Mohamedi, 2017). Under increasing scan rate up to 100 mV s^-1, the Fe₂N@hCC electrodes show over 5-times higher specific capacity than their Fe₂O₃ precursors.

The electrochemically active surface areas of the samples do not directly correlate with material activities as supercapacitor electrodes (Supplementary Figure S2). In general, the areas of electrodes fabricated using electrodeposition are higher than those of the samples prepared by the hydrothermal technique. Furthermore, as can be seen the nitridation of electrodeposited film results in decrease of electrochemically active surface while the effect is reverse for hydrothermal materials. This is consistent with the SEM images that show material shrinking in case of electrodeposited electrode and particles growth in case of hydrothermally obtained films. The limited electrochemically active surface area seen in the Fe₂O₃@hCC-HT electrode may be attributed to the nanoscale dimensions of its particles, which could potentially lead to a lack of electrical contact between some nanoparticles and the hCC electrode.

EIS measurements were performed to gain further insight into the electrochemical characteristics of the synthesized materials, and the results are presented in the form of a Nyquist plots (Figure 7B). Supplementary Figure S3 presents the fitted equivalet circuit which consists of bulk resistance (R_b), a constant phase elements (CPE), resistance related to surface porosity (R_p), a charge transfer resistance (R_ct) and the Warburg impedance (Wu et al., 2021). The joint electrolyte resistance and current collector resistance defined as bulk resistance (R_b), determined from the intersection of the Z′ were 0.09, 0.18 and 0.25 Ω for Fe₂N@hCC-HT, Fe₂N@hCC-E and both Fe₂O₃@hCC, respectively (Gajewska et al., 2023). The R_p values were 0.23, 0.19, 0.61 and 58.4 Ω for Fe₂N@hCC-HT, Fe₂N@hCC-E, Fe₂O₃@hCC-E and Fe₂O₃@hCC-HT, respectively, suggesting improved conductivity within iron-based films obtained from electrodeposition technique. No distinct semicircle in the high-frequency range of the Nyquist plots was observed, which suggests minimized resistance between the binder-free electrode made of iron compounds deposited at the hCC and a current collector (Allison and Andreas, 2019). The charge transfer resistance R_ct was significantly lower for iron nitrides than their Fe₂O₃ counterparts, which implies that nitrides exhibit much faster electron transfer during the electrochemical reaction (Fe₂N@hCC-HT (0.22 Ω) vs. Fe₂O₃@hCC-HT (0.41 Ω) and Fe₂N@hCC-E (0.05 Ω) vs. Fe₂O₃@hCC-_E (0.21 Ω)) (Jiang et al., 2016). Next, a straight 45° line indicates the presence of Warburg impedance, which corresponds to diffuse layer resistance, and then in the low-frequency region, the more inclined the plot is towards the -Z″ axis, the lower the resistance related to the equilibrium differential capacitance. Among all tested samples, Fe₂N@hCC-E exhibits the most vertical line, confirming the formation of an electric double layer at the electrode/electrolyte interface, further indicating a good electric charge storage performance (Mei et al., 2018; Zhao et al., 2023). The EIS data is in good agreement with the CV and GCD measurements, showing that Fe₂N exhibits better electrochemical performance than Fe₂O₃, an additionally confirming that the electrodeposition method for iron nitride precursor is more efficient and promising approach than hydrothermal and drop casting methods.

The long-term performance of the electrode material is another factor in the evaluation of its applicability in energy storage devices. The results of the cyclic stability tests for 2000 cycles at a current density of 1 A g^-1 are presented in Supplementary Figure S4. While the long-term performance of the Fe₂N@hCC electrodes from both preparation methods is very close, the Fe₂O₃-based electrodes show significant discrepancies, emphasizing the superiority of the electrodeposition method over the hydrothermal/drop casting technique. The highest capacity retention was observed for Fe₂O₃@hCC-E (65%), followed by Fe₂N@hCC-E (61%), Fe₂N@hCC-E (56%), and Fe₂O₃@hCC-HT (36%). The trend is similar to the results of the electrochemically active surface area, where materials for electrodeposition exhibit a higher surface accessible for EDL charge storage, resulting in slower electrochemical dissolution of Fe-based compounds (Wang et al., 2021).

This work presents a study on the influence of the synthesis protocol of Fe₂O₃ films deposited on carbon cloth, which are the precursors for Fe₂N on carbon cloth binder-free electrodes. The results of the as-prepared iron-based electrodes show that electrodeposition and hydrothermal/drop casting techniques both before and after nitridation reveal distinct differences between the two methods. Hydrothermal synthesis yields highly crystalline α-Fe₂O₃, while XRD diffraction patterns show that from the direct electrodeposition, amorphous Fe₂O₃ interconnected nanosheets are present on the surface of a carbon cloth. XPS studies revealed a superficial oxidation of the Fe₂N to Fe₂O₃ due to the exposure to air. The study demonstrates that both iron oxide and iron nitride from the electrodeposition synthesis exhibit different and superior electrochemical behavior and performance compared to their hydrothermal counterparts. The Fe₂N@hCC-E sample exhibits the highest specific capacity at a scan rate of 1 mV s⁻¹, with a value of 186 mAh g⁻¹. The obtained results show promising potential for the optimization and application of Fe₂N@hCC from combined electrodeposition and nitridation protocols as a binder-free negative electrode for hybrid supercapacitors.