Electrochemical Performance of Iron-Doped Cobalt Oxide Hierarchical Nanostructure

: In this study, hydrothermally produced Fe-doped Co 3 O 4 nanostructured particles are investigated as electrocatalysts for the water-splitting process and electrode materials for supercapacitor devices. The results of the experiments demonstrated that the surface area, speciﬁc capacitance, and electrochemical performance of Co 3 O 4 are all inﬂuenced by Fe 3+ content. The Fe x Co 3-x O 4 with x = 1 sample exhibits a higher BET surface (87.45 m 2 /g) than that of the pristine Co 3 O 4 (59.4 m 2 /g). Electrochemical measurements of the electrode carried out in 3 M KOH reveal a high speciﬁc capacitance of 153 F/g at a current density of 1 A/g for x = 0.6 and 684 F/g at a 2 mV/s scan rate for x = 1.0 samples. In terms of electrocatalytic performance, the electrode (x = 1.0) displayed a low overpotential of 266 mV (at a current density of 10 mA/cm 2 ) along with 52 mV/dec Tafel slopes in the oxygen evolution reaction. Additionally, the overpotential of 132 mV (at a current density of 10 mA/cm 2 ) and 109 mV with 52 mV/dec Tafel slope were obtained for x = 0.6 sample towards hydrogen evolution reaction (HER). According to electrochemical impedance spectroscopy (EIS) measurements and the density functional theory (DFT) study, the addition of Fe 3+ increased the conductivity at the electrode–electrolyte interface, which substantially impacted the high activity of the iron-doped cobalt oxide. The electrochemical results revealed that the mesoporous Fe-doped Co 3 O 4 nanostructure could be used as potential electrode material in the high-performance electrochemical capacitor and water-splitting catalysts.


Introduction
The rapidly growing population and unstructured industrialization pose a serious threat to future energy demand and environmental health. Conventional energy resources are on the verge of their demise. Hence, significant progress is being made in discovering reliable renewable energy resources and their storage technology. The conversion of hydrogen fuel into electrical energy and storing of this electrical energy in supercapacitors could be the best solution to the energy crisis and environmental remediation. Recent advancements have shown that splitting water into hydrogen and oxygen requires an amount of energy that is not economical, and hence, the search to find efficient catalysts to split water at low energy intake and store hydrogen in them is ongoing [1,2]. Photo-and electro-water splitting has gained attention due to the possibility of utilizing solar radiation to harvest clean energy in the form of hydrogen, which can be sorted and used as a clean alternative fuel [3][4][5]. Water oxidation is challenging, as it requires a four-electron transfer and oxygen-oxygen bond formation. Rh-and Ir-based catalysts show good water-oxidation

Synthesis
The high purity nitrate salts Fe(NO 3 ) 2 .6H 2 O and Co(NO 3 ) 2 .6H 2 O were purchased from Sigma Aldrich (St. Louis, MO, USA). The nitrate salts were dissolved in 35 mL of water with 4 gm of urea (CO(NH 2 ) 2 ) and stirred for 30 min. The solution was then transferred to a 45 mL Teflon-lined autoclave and maintained at 120 • C for 17 h, followed by a natural cooling to room temperature [33]. The precipitates were washed with distilled water and acetone several times by centrifugation and then dried for 12 h at 80 • C under vacuum. After this, the derived metal hydroxide composite particles were calcined at 350 • C in the air for 2 h. The schematic of the experimental process is shown in Figure 1 and Table 1 contains the stoichiometry of the salts needed to make Fe x Co 3-x O 4 . transferred to a 45 mL Teflon-lined autoclave and maintained at 120 °C for 17 h, followed by a natural cooling to room temperature [33]. The precipitates were washed with distilled water and acetone several times by centrifugation and then dried for 12 h at 80 °C under vacuum. After this, the derived metal hydroxide composite particles were calcined at 350 °C in the air for 2 h. The schematic of the experimental process is shown in Figure 1 and Table 1 contains the stoichiometry of the salts needed to make FexCo3-xO4.

Characterization:
X-ray diffraction (XRD) was used to examine the crystalline structure of calcined powder using a Bruker D8 Advance X-ray diffractometer (Bruker, Madison, WI, USA). The surface morphology of prepared samples was studied by a scanning electron microscope (SEM) (Phenom 10 KeV, Pleasanton, CA, USA). Autosorb (Quantachrome, model No. AS1MP, Boynton Beach, FL) was used to quantify specific surface area and pore volume at 77K using nitrogen as the adsorbing gas. The specific surface area of as-prepared samples was measured using the Brunauer-Emmett-Teller (BET) method. The Barret-Joyner-Halenda (BJH) model was also used to calculate surface area, average pore volume, and pore radius. FTIR spectra were collected via the Thermo-Fisher Scientific FTIR spectrometer (Nicolet iS10, Thermo Fisher Scientific, Waltham, MA, USA) between 400 and 1000 cm −1 .

Characterization
X-ray diffraction (XRD) was used to examine the crystalline structure of calcined powder using a Bruker D8 Advance X-ray diffractometer (Bruker, Madison, WI, USA). The surface morphology of prepared samples was studied by a scanning electron microscope (SEM) (Phenom 10 KeV, Pleasanton, CA, USA). Autosorb (Quantachrome, model No. AS1MP, Boynton Beach, FL) was used to quantify specific surface area and pore volume at 77K using nitrogen as the adsorbing gas. The specific surface area of as-prepared samples was measured using the Brunauer-Emmett-Teller (BET) method. The Barret-Joyner-Halenda (BJH) model was also used to calculate surface area, average pore volume, and pore radius. FTIR spectra were collected via the Thermo-Fisher Scientific FTIR spectrometer (Nicolet iS10, Thermo Fisher Scientific, Waltham, MA, USA) between 400 and 1000 cm −1 .
The working electrode was prepared with nickel foam. Nickel foam was first cleaned with acetone and 3M HCl solution for 10 min. This process helps to remove the oxide layer from the surface of the nickel foam. After washing and drying nickel foam, the working electrode was made by mixing 80 weight percent of the synthesized powder, 10 weight percent of the acetylene black, and 10 weight percent of the polyvinylidene difluoride (PVDF) in the presence of N-methyl pyrrolidinone (NMP). The slurry paste was obtained after thoroughly mixing the components and was pasted onto the nickel foam. The prepared electrode on nickel foam was placed in a vacuum oven at 60 • C for 10 h. Using an analytical balance (MS105DU, max. 120 g, 0.01 mg of resolution Mettler Toledo, Columbus, OH, USA), the loading on nickel foam was acquired by measuring the nickel foam before and after loading [34]. An electrochemical cell with a platinum wire as a counter electrode, saturated calomel electrode (SCE) as a reference electrode, and pasted samples onto nickel foam as a working electrode were used. The electrode's electrochemical performance was tested in 3M KOH as an electrolyte with a potential range of 0 to 0.6 V. Cyclic voltammetry (CV), galvanostatic-charge-discharge (GCD) techniques, and electrochemical impedance spectroscopy (EIS) were measured using a Versastat4-500 Processes 2021, 9, 2176 4 of 22 electrochemical workstation (Princeton Applied Research, Oak Ridge, TN, USA). Electrochemical impedance spectroscopy (EIS) and chronoamperometry (CA) were performed at 0.6 and 0.55 V (V, SCE), respectively. The reference electrode was calibrated against a reversible hydrogen electrode (RHE) in the 1M KOH electrolyte at room temperature for water splitting. For every measurement, all potentials were converted and referred to the RHE and iR correction. Linear sweep voltammetry (LSV) measurements were performed at 10 mV s -1 for the OER and HER.

Theoretical Study
First-principles density functional theory (DFT) calculations are performed under the Perdew-Burke-Ernzerhof (PBE) [35] type of exchange-correlation functional for selfconsistent calculations and geometry optimizations. Analyses were performed using VASP [36] (Vienna ab initio simulation package) under the projected augmented wave (PAW) [37] type pseudopotential and the plane-wave basis set with the energy cutoff of 400 eV, which is enough for the self-consistent calculation. The electronic density of states was obtained via the advanced hybrid DFT method (HSE06), where the exchange and correlation terms are described by hybrid functional [38] containing 25% Hartree-Fock exchange. A set of Gamma-centered K-points sampled at 5 × 5 × 3 are used for the integration of the Brillouin zone. The convergence criteria for the electronic and ionic selfconsistent calculations were set to be 10 −4 eV and 10 −3 eV, respectively. The Fe x Co 3-x O 4 structure with the variable x ranging from 0 to 2.0 with a step of 0.5 is considered in the calculation, and the associated results are analyzed.  [39]. Due to impurities, there are no diffraction peaks indicating that Fe 3+ has completely dissolved into the Fe x Co 3-x O 4 structure. The lattice parameter of Fe x Co 3-x O 4 derived using TOPAS software (Bruker) is listed in Table 2. The average crystallite size of Co 3 O 4 calculated using Scherrer's formula [40] is listed in Table 2. The average crystallite size of Fe x Co 3-x O 4 falls within the range of 16.91 nm for x = 0.2 to 20.07 nm for x = 1.0. The noticeable increase in the crystallite size is due to bigger Fe 3+ (r ionic~0 .64 Å) replacing Co 3+ (r ionic~0 .61 Å). Moreover, sharp diffraction peaks indicated the crystalline nature of the calcined Fe x Co 3-x O 4 nanostructures.

Structure and Morphology
The FTIR spectrum Figure 2b displays two distinct bands at 538.060 (ν 1 ) and 653.295 (ν 2 ) cm −1 , which arise from the stretching vibrations of the metal-oxygen bonds [41]. The presence of these vibration bands confirms the development of pure phase spinel Fe x Co 3-x O 4 nanostructures. Fe doping shows a shift of stretching peak towards the right with an increase in Fe 3+ content. As the vibrational frequency is inversely proportional to the mass, the increase in vibrational frequency with Fe 3+ is expected due to the lower atomic weight of Fe (55.8 u) replacing the Co (58.9 u) ion.
The N 2 adsorption/desorption was measured at 77 K with a relative pressure P/Po ranging from 0.029 to 0.99. From these adsorption/desorption curves, BET specific surface areas and corresponding BJH pore sizes were calculated for all Fe x Co 3-x O 4 (0 ≤ x ≤ 1.0), as shown in Figure 2c,d. These curves show the most significant number of pores distribution around 2-4 nm for all Fe x Co 3-x O 4 . This pore size distribution range could help boost the diffusion kinetics inside the electrode material [42]. Table 2 lists measured BJH pore radius, volume, and BET surface area of Fe x Co 3-x O 4 samples. The reported higher surface area, in the range of 59.47 to 87.45 m 2 /g, of samples indicates that the hydrated electrolyte ions could have high contact area at the electrolyte/electrode surface for the Faradaic redox reaction [27]. The observed increase in the surface area of Fe x Co 3-x O 4 samples could be due to particle morphology and crystal packing.
face areas and corresponding BJH pore sizes were calculated for all FexCo3-xO4 (0 ≤ x ≤ 1.0), as shown in Figure 2c,d. These curves show the most significant number of pores distribution around 2-4 nm for all FexCo3-xO4. This pore size distribution range could help boost the diffusion kinetics inside the electrode material [42]. Table 2 lists measured BJH pore radius, volume, and BET surface area of FexCo3-xO4 samples. The reported higher surface area, in the range of 59.47 to 87.45 m 2 /g, of samples indicates that the hydrated electrolyte ions could have high contact area at the electrolyte/electrode surface for the Faradaic redox reaction [27]. The observed increase in the surface area of FexCo3-xO4 samples could be due to particle morphology and crystal packing.     Figure 3 shows the SEM images of Fe x Co 3-x O 4 as a function of Fe 3+ content. Fe x Co 3-x O 4 morphology evolves from thin nanosheet-like architecture for x = 0.0 to nanoflower for x = 0.4 to thicker nanoplates for x = 1.0. It is evident that the morphology of Fe x Co 3-x O 4 has a dependence on the Fe 3+ content in the sample. The initial nanosheet-like structure slowly merges and morphs into a thick plate-like structure with increasing Fe 3+ content. Usually, impurity addition influences the size and morphology of a given crystal by participating in the nucleation and growth process, in which several factors integrate to dominate the process. The crystal growth may vary between fractal aggregation in the initial period and subsequent diffusion processes [43,44]. In addition, the morphology would be affected by the nucleation and the growth process, which, in turn, depends on the charge status of the surface states and dangling bonds. Further, the hydrolysis of urea is accompanied by gas formation, which increases the pressure in the system and could perturb nanocrystalline growth and bring morphological changes.

Electrocapacitive Study
The electrochemical performance of Fe x Co 3-x O 4 nanostructures is determined using cyclic voltammetry (CV) and galvanostatic charge-discharge (GCD) techniques. Figure 4a-f) illustrates the cyclic voltammograms acquired at various scan rates (2-300 mV/s) in a voltage window of 0-0.6 V (vs. SCE). The Equation below gives the faradic reaction for Fe x Co 3-x O 4 [32].
peak current (Ipc) both increase. More precisely, both Ipa and Ipc vary linearly with the square root of the scan rate (ν 1/2 ) described by the Randles-Sevcik equation [48]. This relationship indicates that the redox reaction at the electrode-electrolyte interface is fast, quasi-reversible, and limited by electrolyte diffusion [49]. The positive shift in the oxidation peak and the negative shift in the reduction peak potential show that the electrode materials have low resistance and have strong electrochemical reversibility [27]. The specific capacitance was calculated from CV curves using the Equation below [29,50,51]: Here, Csp is the specific capacitance (F/g), I (A) is the charge-discharge current, ΔV (V) is the potential range, m (g) is the mass of the electroactive materials, and t (s) is the discharging time. The Csp of FexCo3-xO4 samples as a function of scan speeds is shown in Figure 5 The total charge storage process is affected by factors such as the faradaic contribution from the insertion process of electrolyte ions, the faradaic contribution from the charge-transfer process, the high surface area contribution from pseudocapacitance, and the non-faradaic contribution from the double layer effects. Data from the CV are  Oxidation and reduction peaks in the cathodic and anodic scans were seen in all CV curves [45]. The pseudocapacitive character of electrodes is indicated by non-rectangular and asymmetric CV curves [46]. The CV curves of Fe x Co 3-x O 4 show no additional redox peaks, indicating that the redox processes in Fe x Co 3-x O 4 are quite similar to those in Co 3 O 4 . The peaks are due to the redox reaction related to M-O/ M-O-OH, where M stands for Fe or Co ions [28]. A positive shift in oxidation peak potential and a negative shift in reduction peak potential are seen when the scan rate increases [47]. This PC characteristic is derived from the faradaic redox reaction related to the reversible reaction of Fe 2+ /Fe 3+ and Co 3+ /Co 4+ transitions. Figure 5a shows the Randles-Sevcik plots of the Fe x Co 3-x O 4 samples. When the scan rate increases from 2 to 300 mV s −1 , the anodic peak current (Ipa) and the cathodic peak current (Ipc) both increase. More precisely, both I pa and I pc vary linearly with the square root of the scan rate (ν 1/2 ) described by the Randles-Sevcik equation [48]. This relationship indicates that the redox reaction at the electrode-electrolyte interface is fast, quasi-reversible, and limited by electrolyte diffusion [49]. The positive shift in the oxidation peak and the negative shift in the reduction peak potential show that the electrode materials have low resistance and have strong electrochemical reversibility [27]. The specific capacitance was calculated from CV curves using the Equation below [29,50,51]: each contribution.
where k1 and k2 are determined from the Csp vs. v −1/2 linear plot; k2 is the slope, and k1 is the intercept. k1 indicates diffusion, and k2 shows capacitance contribution to the total specific capacitance for a given voltage. For the calculation, the specific capacitance was plotted against the slow scan rate up to a value of 20 mV/s and a regression fit was performed using Equation (3). The obtained values for k1 and k2 were used to determine the fractional contribution in terms of diffusion and capacitance from the total specific capacitance, given in Figure 5c. Figure 5c shows that the contribution to the current response at a fixed potential is more capacitive than diffusive, increasing with the Fe content.
Galvanostatic charge-discharge (GCD) experiments were carried out in 3M KOH solution within the voltage window of 0.0 to 0.6 at varied current densities ranging from 1 A/g to 30 A/g to further quantify the potential applicability of as-prepared electrodes for supercapacitors. Figure 6a-f shows GCD curves. These results indicate that FexCo3-xO4 electrodes can charge and discharge quickly with good electrochemical reversibility at various constant current densities. The non-linear relationship between potential and time in both charge-discharge cycles, as in CV curves, indicates that the capacitance of the studied nanomaterials is not constant between potential ranges and reflects typical PC behavior [58]. The non-linearity of the GCD curve is a consequence of the Co 3+ /Co 4+ ions' redox reactions with OH -. The discharge process is divided into three distinct steps: the first is a rapid potential drop caused by internal resistance; the second is a slow potential decay at an intermediate time caused by the faradaic redox reaction, and the third is a fast potential decay caused by electric double layer capacitance [30]. The specific capacitance for electrodes of FexCo3-xO4 were calculated using Equation (1).  The total charge storage process is affected by factors such as the faradaic contribution from the insertion process of electrolyte ions, the faradaic contribution from the chargetransfer process, the high surface area contribution from pseudocapacitance, and the non-faradaic contribution from the double layer effects. Data from the CV are evaluated at various scan rates and can be used to characterize capacitive effects by equation [52].
Here, the peak current is I (A), the scan rate is v (mV/s), and a and b are adjustable parameters. The value of b defines the charge storage mechanism and b-values as a function of potential for the cathodic sweeps. The parameter b value is determined by plotting log(i) versus log(v) curves. If b = 1, then the capacitive surface mechanism is dominant, which indicates a fast near-surface controlled reaction, and if b = 0.5, it indicates a diffusioncontrolled faradaic reaction during the charge storage mechanism [53]. Figure 5c shows the peak current vs. square root of the scan rate curve. These curves are fitted with Equation (2), which gives values of b~0.5392, 0.4792, 0.3714, 0.455, 0.4336, and 0.4158 for x = 0.0, 0.2, 0.4, 0.6, 0.8, and 1.0, respectively. When the b value approaches 0.5, it indicates that the electrochemical reaction is dominated by ionic diffusion, whereas, when it approaches 1, the capacitive behavior dominates the total process [54]. Furthermore, the lower value of the scan rate region in Figure 5c exhibits a linear relationship with the square root of the scan rate, indicating a quasi-reversible electrochemical process [55]. Figure 5c shows that, at scan rates greater than 100 mV/s, ion diffusion is limited to the surface of the active material of the electrode, i.e., EDLC dominates over the pseudocapacitor, and OH − diffusion can adhere only to the outer layer of the nanostructure, which contributes less to the electrochemical capacitive behavior [56]. On the other hand, at scan rates lower than 100 mV/s, the faradaic redox reaction dominates due to more efficient use of the active material in the working electrode, and diffusion of OH − ions can easily penetrate deep into the nanostructure's interlayer, resulting in the adsorption of more ions and thus resulting in a higher specific capacitance [56].
There are two different mechanisms for current response at a fixed potential: the first is capacitive, and the second is the diffusion process [57]; the Equation below gives each contribution.
where k 1 and k 2 are determined from the C sp vs. v −1/2 linear plot; k 2 is the slope, and k 1 is the intercept. k 1 indicates diffusion, and k 2 shows capacitance contribution to the total specific capacitance for a given voltage. For the calculation, the specific capacitance was plotted against the slow scan rate up to a value of 20 mV/s and a regression fit was performed using Equation (3). The obtained values for k 1 and k 2 were used to determine the fractional contribution in terms of diffusion and capacitance from the total specific capacitance, given in Figure 5c. Figure 5c shows that the contribution to the current response at a fixed potential is more capacitive than diffusive, increasing with the Fe content. Galvanostatic charge-discharge (GCD) experiments were carried out in 3M KOH solution within the voltage window of 0.0 to 0.6 at varied current densities ranging from 1 A/g to 30 A/g to further quantify the potential applicability of as-prepared electrodes for supercapacitors. Figure 6a-f shows GCD curves. These results indicate that Fe x Co 3-x O 4 electrodes can charge and discharge quickly with good electrochemical reversibility at various constant current densities. The non-linear relationship between potential and time in both charge-discharge cycles, as in CV curves, indicates that the capacitance of the studied nanomaterials is not constant between potential ranges and reflects typical PC behavior [58]. The non-linearity of the GCD curve is a consequence of the Co 3+ /Co 4+ ions' redox reactions with OH − . The discharge process is divided into three distinct steps: the first is a rapid potential drop caused by internal resistance; the second is a slow potential decay at an intermediate time caused by the faradaic redox reaction, and the third is a fast potential decay caused by electric double layer capacitance [30]. The specific capacitance for electrodes of Fe x Co 3-x O 4 were calculated using Equation (1). The occurrence of a voltage plateau in GCD curves confirms the pseudocapacitance behavior of electrodes concerning their discharging period [31]. The specific capacitance values of FexCo3-xO4, calculated using Equation (1) at 1 A/g, are 131, 146, 131, 153, 137, and 149 F/g for x=0.0, 0.2, 0.4, 0.6, 0.8, and 1.0, respectively. Figure 7a shows the dependence of current density on specific capacitance. A decrease in the specific capacitance with the increase in the discharge current is caused by the insufficient time available for the diffusion of electrolyte ions into the inner electrode surface and the increase in potential drop towards higher discharge currents [59]. The Ragone plot shown in Figure 7b represents energy density vs. power density. The values of energy density and power den- The occurrence of a voltage plateau in GCD curves confirms the pseudocapacitance behavior of electrodes concerning their discharging period [31]. The specific capacitance values of Fe x Co 3-x O 4 , calculated using Equation (1) at 1 A/g, are 131, 146, 131, 153, 137, and 149 F/g for x = 0.0, 0.2, 0.4, 0.6, 0.8, and 1.0, respectively. Figure 7a shows the dependence of current density on specific capacitance. A decrease in the specific capacitance with the increase in the discharge current is caused by the insufficient time available for the diffusion of electrolyte ions into the inner electrode surface and the increase in potential drop towards higher discharge currents [59]. The Ragone plot shown in Figure 7b represents energy density vs. power density. The values of energy density and power density were calculated from GCD data according to the following equations,

(c)
where E g (Wh/kg) is the average energy density, C sp (F/g) is the specific capacitance, and dV (V) is the potential difference drop during the discharging process. P g (W/kg) is the average power density, and dt(s) is the discharge time [60]. The specific capacitance, stability, and power density of the Fe x Co 3-x O 4 electrode are compared with other reported metal oxides in Table 3. According to the results, the proposed catalyst has comparable or higher power density, stability, and specific capacitances to many other metal oxide electrocatalysts.  The occurrence of a voltage plateau in GCD curves confirms the pseudocapacitance behavior of electrodes concerning their discharging period [31]. The specific capacitance values of FexCo3-xO4, calculated using Equation (1) at 1 A/g, are 131, 146, 131, 153, 137, and 149 F/g for x=0.0, 0.2, 0.4, 0.6, 0.8, and 1.0, respectively. Figure 7a shows the dependence of current density on specific capacitance. A decrease in the specific capacitance with the increase in the discharge current is caused by the insufficient time available for the diffusion of electrolyte ions into the inner electrode surface and the increase in potential drop towards higher discharge currents [59]. The Ragone plot shown in Figure 7b represents energy density vs. power density. The values of energy density and power density were calculated from GCD data according to the following equations, (4) Where Eg(Wh/kg) is the average energy density, Csp (F /g) is the specific capacitance, and dV (V) is the potential difference drop during the discharging process. Pg (W/ kg) is the average power density, and dt(s) is the discharge time [60]. The specific capacitance, stability, and power density of the FexCo3-xO4 electrode are compared with other reported metal oxides in Table 3. According to the results, the proposed catalyst has comparable or higher power density, stability, and specific capacitances to many other metal oxide electrocatalysts.   The cyclic stability of Fe x Co 3-x O 4 electrodes was evaluated by repeated chargedischarge measurements up to 5000 cycles at a constant current density of 10 A/g in the potential range between 0.0 to 0.6 V in 3M KOH, as shown in Figure 8. The percentage retention in specific capacitance was calculated using Equation (6): % Retention in specific capacitance = (C # /C 1 ) × 100 (6) where C # is specific capacitance at various cycles, and C 1 is the specific capacitance at the first cycle. Hence, the % retention in specific capacitance after 5000 cycles x = 0.0, 0. The cyclic stability of FexCo3-xO4 electrodes was evaluated by repeated charge-discharge measurements up to 5000 cycles at a constant current density of 10 A/g in the potential range between 0.0 to 0.6 V in 3M KOH, as shown in Figure 8. The percentage retention in specific capacitance was calculated using Equation (6): % Retention in specific capacitance = (C#/C1)×100 (6) where C# is specific capacitance at various cycles, and C1 is the specific capacitance at the first cycle. Hence, the % retention in specific capacitance after 5000 cycles x = 0.0, 0.2, 0.4, 0.6, 0.8, and 1.0 are 80.4%, 91.3%, 63.4%, 76.7%, 77.4%, and 76.1%, respectively. This shows that, even after 5000 cycles, the electrode shows outstanding cyclic stability, and hence, there is not so much difference in the specific capacitance from their initial values in x = 0.0. Furthermore, the Coulombic efficiency for x=0.0, 0.2, 0.4, 0.6, 0.8, and 1.0 after 5000 cycles is 100%, 98.5 %, 96.2%, 94.5%, 94.4%, and 93%, respectively. Both % retention and Coulombic efficiency are shown in Figure 8a-f, respectively.

DFT Study
In our DFT computation, we are dealing with a cell with 28 atoms, 12 Co atoms, and 16 oxygen atoms. There are eight octahedra and four tetrahedra sites in Co. So, our first

DFT Study
In our DFT computation, we are dealing with a cell with 28 atoms, 12 Co atoms, and 16 oxygen atoms. There are eight octahedra and four tetrahedra sites in Co. So, our first job was to find the most favorable sites for Fe. Calculations show that Fe atoms first prefer to go to octahedral sites. After all of them are occupied, they go to tetrahedra sites. It is also found that the Fe atoms occupying two different sites prefer to have opposite magnetic moments. Those that occupy octahedral sites have negative magnetic moments, and tetrahedral sites have positive magnetic moments. Figure 2d shows that the bandgap varies from 0.60 eV to 1.09 eV for the different values of x from 0.0 to 2.0. For the pure Co 3 O 4 system, a value of 1.09 eV is obtained, which is close to the reported value under the GW approximation [69]. Doping of Fe over Co causes the bandgap to decrease, reaching a minimum, and then increase. The smallest bandgap has a value of 0.60 eV corresponding to the FeCo 2 O 4 compound. The calculation thus shows that FeCo 2 O 4 could be more conductive than other forms of Fe x Co 3-x O 4 (0 ≤ x ≤ 2), which is important for practical applications.

Electrocatalytic Behavior
The electrochemical water splitting performance of Fe x Co 3-x O 4 samples as an electrocatalyst was explored in a 1 M KOH electrolyte. Firstly, the OER activity of all samples was analyzed using linear sweep voltammetric (LSV) at a scan rate of 2 mV/s in Figure 9a. It was observed that the introduction of Fe 3+ induced a change in electrocatalytic behavior by leading to an improvement in electrocatalyst performance. As a result, the onset overpotential of 245 mV and overpotential of 266 mV was achieved for the x = 1.0 sample to deliver the current density of 2 and 10 mA/cm 2 , respectively. Table 4 shows the onset overpotential (at 2 mA/cm 2 current density) and overpotential (at 10 mA/cm 2 current density), both of which are indicators of electrocatalytic activity. Tafel slopes are calculated using the following Equation: η = a + b log j, where η is the overpotential, a is the constant, b is the Tafel slope, and j is the current density. Figure 9b  to deliver the current density of 2 and 10 mA/cm 2 , respectively. Table 4 shows the onset overpotential (at 2 mA/cm 2 current density) and overpotential (at 10 mA/cm 2 current density), both of which are indicators of electrocatalytic activity. Tafel slopes are calculated using the following Equation: η = a + b log j, where η is the overpotential, a is the constant, b is the Tafel slope, and j is the current density. Figure     The stability of the electrocatalyst is also a crucial factor for catalyst development, so the first LSV polarization curve was compared with the curve after CV 1000 cycles. Figure 10 showed almost perfect overlaps of the 1stand 1000th cycles, suggesting all electrocatalysts have great stability towards OER. The electrocatalytic HER performance was also investigated using LSV at 2 mV/s in Figure 11a. After being Fe 3+ doped (x = 0.6), cobalt oxide with an overpotential of 182 mV showed a greatly reduced overpotential of 136 mV (at the 10 mA/cm 2 current density) and a low onset overpotential of 1 mV (at the 2 mA/cm 2 current density). A tendency to improve HER performance was observed after Fe's introduction. The onset overpotential and overpotential are listed in Table 4. Figure 11b shows the Tafel slope for all samples, and the observed Tafel slope was 97, 116, 109, 109, 114, and 108 for x = 0.0, 0.2, 0.4, 0.6, 0.8, and 1.0, respectively. For stability towards HER, the LSV 1st graph and 1000th graph were compared, as seen in Figure 12. There was no discernible difference when comparing the two curves, demonstrating that the electrodes are quite stable for HER. For both the HER and OER processes, the acquired results reveal an adequate and long-lasting response. As a result, the given results make the material appealing for practical industrial electrolysis applications. Figure 13 displays the Nyquist plots for all samples. Electrochemical impedance spectroscopy (EIS) measurements were performed at 10 mV AC amplitude in the frequency range of 0.05 Hz to 10 kHz. The electrolyte resistance is represented by the intersection value of the real axis, which is roughly 0.5/cm 2 for all samples. This resistance is generally attributed to the wire, electrode material, and electrolyte resistance [70]. In addition, the diameter of the semicircle indicates the resistance at the interface between the Fe x Co 3-x O 4 electrodes and the 1 M KOH electrolyte. All graphs showed a decrease in the diameter with increased voltage and decreased resistance in the order of x = 1.0, 0.2, 0.8, 0.4, 0.6, and 0.0. The introduction of Fe 3+ increased the conductivity at the interface between the electrode and the electrolyte, which significantly influenced the high activity of the iron-doped cobalt oxide.
Therefore, the impedance data were analyzed using a modified version of the wellknown equivalent electric circuit of Armstrong and Henderson [71,72]. For example, Figure 14 presents the experimental and simulated EIS spectra of Fe x Co 3-x O 4 at x = 0.0 and x = 1.0, along with their equivalent circuits. All samples show a similar circuit behavior. The figures represent experimental data (exp), and the solid line represents the simulation data (sim). The equivalent circuits consist of electrolyte resistance R s , solid electrolyte interphase resistance R f , solid electrolyte interphase capacitance C f , charge-transfer resistance R ct , double layer capacitance C dl , and Warburg diffusion impedance W, respectively. The elements of the equivalent circuit help understand the charge transfer process during electrochemical analysis [73]. EIS curves begin with a suppressed semicircle at a low-frequency range followed by another highly eccentric arc at a higher frequency region. This arc may be related to electrode-electrolyte interphase resistance coupled with double-layered capacitance [74].
The simulated values of resistance R s of the sample Fe x Co 3-x O 4 are shown decreasing from x = 0.0 to x = 1.0, as can be realized by looking at the shifted curves of Figure 14a,b towards the lower value of the real impedance Z' axis. The decreasing value of R s and R f may be due to the higher content of Fe in the sample and increased grain boundaries to make a smooth electrode-electrolyte interface. Such an interface allows a more ionic path to perform an excellent electrochemical reaction [75]. Furthermore, the Fe x Co 3-x O 4 samples showed a low loss of current density during prolonged chronoamperometry experiments, Figure 15, indicating that the electrode has the potential to maintain stability over time.
Although there is some fluctuation owing to gas generation, all electrodes have excellent endurance, since the graph remains stable for a long time.  Figure 13 displays the Nyquist plots for all samples. Electrochemical impedance spectroscopy (EIS) measurements were performed at 10 mV AC amplitude in the frequency range of 0.05 Hz to 10 kHz. The electrolyte resistance is represented by the intersection value of the real axis, which is roughly 0.5/cm 2 for all samples. This resistance electrode-electrolyte interface. Such an interface allows a more ionic path to perform an excellent electrochemical reaction [75]. Furthermore, the FexCo3-xO4 samples showed a low loss of current density during prolonged chronoamperometry experiments, Figure  15, indicating that the electrode has the potential to maintain stability over time. Although there is some fluctuation owing to gas generation, all electrodes have excellent endurance, since the graph remains stable for a long time.  Table 5 compares the OER/HER performance of the FexCo3-xO4 sample (x = 1.0) to that of other doped cobalt-based materials previously reported. Our findings demonstrate indisputably that the FexCo3-xO4 compound is an effective electrocatalyst for OER/HER applications. Several reasons should be examined to explain improved electrocatalytic performance after Fe 3+ is doped in cobalt oxide. The expansion of the active sites due to the increase in surface area and the increase in conductivity due to the decrease in the bandgap could have influenced the electrocatalysis activity. In addition, a  Table 5 compares the OER/HER performance of the FexCo 3-x O 4 sample (x = 1.0) to that of other doped cobalt-based materials previously reported. Our findings demonstrate indisputably that the FexCo 3-x O 4 compound is an effective electrocatalyst for OER/HER applications. Several reasons should be examined to explain improved electrocatalytic performance after Fe 3+ is doped in cobalt oxide. The expansion of the active sites due to the increase in surface area and the increase in conductivity due to the decrease in the bandgap could have influenced the electrocatalysis activity. In addition, a preliminary DFT study looking at the adsorption energy showed that incorporating Fe atoms into the lattice of Co 3 O 4 significantly lowers the adsorption energy difference from OH* to O* from 1.90 to 1.54 eV [76]. Given the hierarchical structure of FexCo 3-x O 4 , all doped Fe 3+ are expected to reside near the surface [77]. Thus, our results show that the proposed catalyst can be capitalized to produce hydrogen and oxygen gas at a considerable rate due to the cost-effective, earth-abundant, and inexpensive catalyst.

Conclusions
In conclusion, a simple and cost-effective strategy using the hydrothermal method has been developed to report the effect of Fe 3+ doping on the structural and electrochemical performance of the Co 3 O 4 nanostructure. The experimental results revealed the dependence of surface area, specific capacitance, bandgap, and electrochemical performance of Co 3 O 4 on Fe 3+ content. Excellent electrochemical performance was obtained for Fe x Co 3-x O 4 , x = 1.0. Moreover, the appropriate doping of Fe atoms has modified the morphology of Co 3 O 4 particles, resulting in the shortening of the ion transportation path, which provided much better conductivity. The electrochemical characterization shows that the nanocomposite shows low dynamic potential compared to its counterparts for both HER and OER in alkaline solution. According to the findings, Fe 3+ can be a cost-effective and good substitute for cobalt in Co 3 O 4 to achieve excellent electrochemical and water-splitting performance.