The rapidly increasing demand for sustainable and renewable clean energy sources has attracted considerable interest in promoting the development of energy storage and conversion technologies with low cost, high efficiency and environmental benignity [1, 2]. Electrochemical water splitting, an efficient and environmentally friendly technology for producing hydrogen fuel and oxygen, is a promising pathway to convert electrical energy into chemical energy. However, the electrolytic efficiency is limited by the oxygen evolution reaction (OER) [3, 4]. To date, IrO₂ and RuO₂ are known as the most active OER catalysts; however, some limitations exist such as high cost and relatively poor stability in alkaline medium. Therefore, it is desirable to explore inexpensive, highly efficient and stable OER catalysts based on earth-abundant elements [5-7].

Over the past decades, earth-abundant transition metal catalysts, particularly Co-based catalysts, including metal oxides [8, 9], oxyhydroxides [10, 11], nitrides [6, 12], phosphides [13, 14], sulfides [15], selenides [16] and perovskite solids [17], have been extensively synthesized and applied in water splitting. However, well-studied cobalt oxides or (oxy) hydroxides usually suffer from low conductivity and reactivity, in comparison to the corresponding multimetal electrocatalysts doped with other metal such as Fe and Ni [4, 11], which results in large overpotentials required to reach the desired current density of 10 mA cm^–2 during the OER process [5, 13]. Therefore, significant efforts have been directed toward improving their catalytic performance [18, 19]. Increasing the number and reactivity of active sites are the two main strategies for enhancing their catalytic performance. The first could be achieved by tuning the morphology and structure, such as nanosheet arrays [4, 20], ultrathin characteristics [21, 22] and porous structure [23-25], to enhance their OER performance. For example, a porous structure is one of the most important requirements because it facilitates the increase in specific surface area, buffering of volume change, and gas release [26, 27]. Jiao et al. [9] synthesized ordered mesoporous cobalt oxide as a highly efficient oxygen evolution catalyst. The second is the incorporation of other components to optimize the electronic structure, thereby increasing the reactivity of the catalysts [5]. Over the last few years, it has been reported that the incorporation of Fe could significantly increase the performance of Co-based catalysts. Han et al. [19] synthesized nanosized porous Fe-CoOOH via etching the CoFeAl-LDH/G in an alkali solution to enhance the activity of Co-based catalysts. These nanosized Fe-CoOOH/G are ultrasensitive and highly effective for OER. Li et al. [23] prepared activating CoOOH porous nanosheet arrays by partial Fe substitution for efficient oxygen evolution reaction. Thus, the development of novel catalysts combing these two strategies for OER is of importance.

Herein, we have successfully synthesized 3D porous FeCo oxyhydroxide layer coated on the carbon cloth (3D-FeCoOOH/CC) as a highly efficient electrode with a facile and cost-effective electro-oxidation method. The Fe components and unique 3D porous structures were successfully introduced and created after Fe doping and in situ electro-oxidation process, which favored the exposure of more active sites, thereby accelerating the diffusion of electrolytes and electronic transfer, as well as effectively buffering the volume change for oxygen release. The electrochemical testing results showed that the as-synthesized 3D-FeCoOOH/CC exhibited a superior OER activity with a relatively low overpotential of 259 mV at 10 mA cm^–2, with a small Tafel slope of 34.9 mV dec^–1, and excellent stability in 1.0 mol L^–1 KOH solution, thereby outperforming commercial IrO₂ catalysts.

Carbon cloth (CC) was purchased from Shanghai Hesen Electric Co., Ltd. Cobalt nitrate hexahydrate (Co(NO₃)₂·6H₂O), hydrochloric acid (HCl), and ethanol were purchased from Aladdin Chemistry Co. Ammonium fluoride (NH₄F), urea, ammonium iron(II) sulfate ((NH₄)₂Fe(SO₄)₂), and potassium hydroxide (KOH) were purchased from Sinopharm Chemical Reagent Co., Ltd. Nafion (5 wt.%) was purchased from Dupont Co. Iridium oxide (IrO₂) was purchased from Makclin Chemistry Co. All chemicals were used as received without any further purification. Distilled water was obtained through the Millipore system.

Typically, MCPAs/CC is prepared by a modified hydrothermal reaction [28]. A piece of CC (2 cm × 4 cm) was sonicated in HCl, acetone and deionized water sequentially for 30 min to ensure that the surface of CC was clean. Co(NO₃)₂·6H₂O (2.5 mmol), NH₄F (2.5 mmol), and urea (12.5 mmol) were dissolved in a mixture of 40 mL ethylene glycol and 5 mL deionized water under vigorous stirring for 30 min to obtain a uniform solution. Subsequently, the uniform solution and a piece of clean CC were transferred into a 100 mL Teflon-lined stainless-steel autoclave and heated at 120 ℃ for 6 h. After the autoclave naturally cooled down to room temperature, the MCPAs/CC was retrieved, washed with ethanol repeatedly, and dried at 60 C for 4 h.

The 3D-FeCoOOH/CC was fabricated via two steps. First, the Fe doping process was carried out via the in situ electro-oxidization of MCPAs/CC in 0.01 mol L^–1 (NH₄)₂Fe(SO₄)₂ at a current density of 1.2 mA cm^–2 for a specific time period. To obtain the optimum catalytic activity, the dopant process time was set at 15, 30 and 45 min. Subsequently, the precursors were cleaned repeatedly using distilled water to remove the excess Fe, followed by direct use as the working electrode in 1.0 mol L^–1 KOH at an anodic current density of 10.0 mA cm^–2 for 10 h to obtain 3D-FeCoOOH/CC. Unless specifically noted, the 3D-FeCoOOH/CC mentioned later in this article refers to the sample with an Fe doping time of 30 min.

For comparison, the 3D-CoOOH/CC was fabricated using the above-mentioned method with the absence of the Fe doping process.

A suspension was prepared by dispersing IrO₂ powders (5 mg) in a mixture of 0.99 mL ethanol and 0.01 mL 5 wt.% Nafion solution. The mixed solution was sonicated for 20 min to obtain a homogeneous catalyst ink. The dispersion (0.2 mL) was transferred onto a piece of CC (1 cm × 2 cm) using a pipette. The mass loading was controlled at 1.0 mg cm^–2.

The catalyst loading was measured as 1.25 mg cm^–2 by the traditional gravimetric method. The morphology and structures were analyzed using field emission scanning electron microscopy (SEM; Hitachi S-4800F) and high-resolution transmission electron microscopy (HRTEM; FEI Tencnai G2F30). The powder X-ray diffraction (XRD) patterns were collected on a Bruker D8AVANCE X-ray powder diffraction spectrometer. Nitrogen adsorption-desorption isotherms were measured using a Micromeritics ASAP 2000 system. The valency of the different components of the catalysts was evaluated by X-ray photoelectron spectroscopy (XPS; VG-Multi-lab 2000).

All electrochemical measurements were carried out in a standard three-electrode system with a CHI 660E electrochemical analyzer (CH Instruments, Inc., Shanghai) using 1.0 mol L^–1 KOH as the electrolyte. The synthesized 3D-FeCoOOH/CC was used as the working electrode, a graphite plate as the counter electrode, and Hg/HgO electrode as the reference electrode. The potentials reported in this study were calibrated to the reversible hydrogen electrode (RHE), using the following equation:

(1)

The electrochemically active surface area (ECSA) was estimated from the electrochemical double-layer capacitance based on a published report [29]. The electrochemical capacitance (C_dl) was estimated by cyclic voltammograms measured in a non-Faradaic region from 1.18 V to 1.26 V at various scan rates of 5–25 mV s^–1. The ECSA was calculated according to Equation (2) below:

(2)

The general specific capacitance for the metal oxides or oxyhydroxides, C_s = 0.040 mF cm^–2 in 1.0 mol L^–1 KOH was employed based on previous reports [29].

The overall synthetic strategy for 3D-FeCoOOH/CC is illustrated in Fig. 1. First, MCPAs/CC was synthesized by a simple hydrothermal reaction and directly used as a working electrode. Second, Fe doping process was carried out via the in situ anodic oxidization of MCPAs/CC in 0.01 mol L^–1 (NH₄)₂Fe(SO₄)₂ for a fixed period of time, at 1.2 mA cm^–2. The electrode was in situ electro-oxidized in 1.0 mol L^–1 KOH solution at 10 mA cm^–2 for 10 h continuously, to afford 3D-FeCoOOH/CC. To obtain the optimum catalytic activity, the Fe doping process was fixed at different periods and the corresponding catalytic activity is shown in Fig. S1. It is evident that the 3D-FeCoOOH/CC with an Fe doping time 30 min in (NH₄)₂Fe(SO₄)₂ possesses the best OER catalytic activity. As 3D-FeCoOOH/CC exhibited the best activity, in-depth characterizations were performed on it.

The morphologies of MCPAs/CC and 3D-FeCoOOH/CC were investigated using SEM. Fig. 2a shows the low-magnification SEM image of MCPAs/CC. It is clearly observed that numerous micro go pieces (~ 1.3 μm) are coated on the carbon fiber. The high-magnification SEM images in Figs. 2b and 2c show that every micro go piece is composed of a large number of uniformly intertwined nanorods. The corresponding energy-dispersive X-ray spectroscopy (EDS) spectrum in Fig. S2 demonstrates that it contains C, O, F and Co. For 3D-FeCoOOH/CC, the low- and high-magnification SEM images in Figs. 2d and 2e show that the carbon fiber is coated with a uniform layer, which comprises of numerous small nanoparticles (~ 10 nm). The corresponding EDS spectrum in Fig. S3 shows the presence of C, O, Fe and Co elements (Fe:Co molar ratio of 0.034:1), indicating the dopant of Fe element and the complete removal of F. Besides, the SEM images and the corresponding EDS spectra of 3D-FeCoOOH/CC with the Fe dopant times of 15 and 45 min are shown in Figs. S4–S7. The SEM results demonstrate that 3D-FeCoOOH/CC has a unique 3D porous structure, and that the doped Fe atoms play a key role in tailoring and converting MCPAs/CC into 3D-FeCoOOH/CC [19]. Electrochemical tests indicate that the 3D-FeCoOOH/CC obtained with an Fe doping time of 30 min exhibits the best catalytic activity, indicating a suitable Fe:Co ratio for OER. For comparison, 3D-CoOOH/CC was synthesized exclusively by the electro-oxidation in 1.0 mol L^–1 KOH without the Fe doping process. The low-magnification SEM image in Fig. 2g shows that 3D-CoOOH/CC still exhibits a micro go piece morphology without obvious changes, whereas the high-magnification SEM images in Figs. 2h and 2i demonstrate that the nanorods are completely transformed into small nanoparticles (~ 10 nm), similar to the morphology of 3D-FeCoOOH/CC.

The nanostructure of the 3D-FeCoOOH layer scraped from carbon fiber was investigated by transmission electron microscopy (TEM). As shown in Fig. 3a, the 3D-FeCoOOH layer has a porous structure. HRTEM and selected area electron diffraction (SAED) analyses were performed to further study the morphology and crystallographic properties. The HRTEM image in Fig. 3b and the inset corresponding to the SAED pattern both indicate that 3D-FeCoOOH/CC exhibits a polycrystalline structure. The first, third and fourth diffraction rings close to the central diffraction spot in the SAED pattern correspond to the crystallographic planes of CoOOH (110), CoOOH (021) and CoOOH (140), respectively. Meanwhile, the second diffraction ring corresponds to the crystallographic plane of FeOOH (310), suggesting that FeOOH and a main CoOOH phase exist. High angle annular dark field scanning TEM (HADDF-STEM) images and the corresponding elemental mappings shown in Figs. 3c–f indicate that Co (red), O (blue) and Fe (yellow) are distributed uniformly. The formation of a porous structure was further defined using Brunauer-Emmett-Teller (BET) gas sorptometry measurements, and the corresponding results are shown in Fig. S8. The shapes of the hysteresis loops of 3D-CoOOH/CC and 3D-FeCoOOH/CC were both identified as type H3 according to the International Union of Pure and Applied Chemistry (IUPAC) classification [4], indicating the formation of porous structures. This result was consistent with the SEM data. Additionally, according to the BET analysis, the total specific surface areas of MCPAs/CC and 3D-CoOOH/CC are 23.59 m² g^–1 and 25.80 m² g^–1, respectively. When Fe was introduced into 3D-CoOOH/CC, the specific surface area of 3D-FeCoOOH/CC further increased to 27.24 m² g^–1, which was facilitated the exposure of more active sites.

The XRD patterns of the as-synthesized electrodes and CC are shown in Fig. S9. Fig. S9a shows two obvious diffraction peaks for 3D-FeCoOOH/CC at 26.4° and 54.5°, which could be attributed to the (002) and (004) planes of CC. From the magnification in Fig. S9b, notably, no obvious diffraction peak for 3D-FeCoOOH/CC was observed. Fourier transform infrared spectroscopy (FT-IR) analysis was further performed to explore the chemical composition of 3D-FeCoOOH/CC. It can be concluded from Fig. S10 that MCPAs/CC is most likely composed of fluorinated ethylene glycol alkoxide, and that the obtained products after the two-step electrochemical oxidation process are mainly metal oxyhydroxides [35].

High-resolution XPS was carried out to determine the valency of different components. The XPS survey of MCPAs/CC presented in Fig. S11 shows the presence of C, O, F and Co elements, while 3D-FeCoOOH/CC shows the presence of C, O, Fe and Co elements. The Co 2p high-resolution XPS spectrum of MCPAs/CC is shown in Fig. S12. The binding energies of Co 2p_1/2 and Co 2p_3/2 are located at 807.8 and 791.9 eV, respectively, revealing the Co²⁺ oxidation state. Fig. 4a shows the comparison of the Co 2p high-resolution XPS spectra for 3D-CoOOH/CC and 3D-FeCoOOH/CC. For 3D-CoOOH/CC, the two peaks with binding energies of 780.5 and 795.7 eV correspond to the typical Co 2p_3/2 and Co 2p_1/2 orbitals, while for 3D-FeCoOOH/CC, these could be attributed to 780.9 and 796.1 eV. This result indicates that Co is present in the trivalent oxidation state [19, 32]. As expected, the high-resolution Co 2p spectra of 3D-FeCoOOH/CC exhibits a 0.4 eV shift to a higher binding energy compared with the trend exhibited for 3D-CoOOH/CC, indicating a modulated electronic structure of the Co atoms and a strong electronic interaction between Co and Fe atoms [19]. Moreover, the enhanced Co satellite peak of 3D-FeCoOOH/CC at ~ 785 eV indicates the rearrangements of Co 3d electron population owing to the degeneracy breaking of Co ions, which is consistent with the reports in the literature [36]. Additionally, the rearrangements of the Co 3d electron population caused an increase in the electron density of the Co atoms, which facilitated the enhancement of catalytic activity [23, 37]. Fig. 4b shows the high-resolution XPS spectrum for Fe 2p of 3D-FeCoOOH/CC. The spectrum displays two major peaks at binding energies of 725.0 eV and 712.5 eV, revealing the characteristics of Fe³⁺ [19, 31]. Furthermore, it has been reported that the adsorption energy of OH is considerably strong for FeOOH, whereas it is too weak on CoOOH [5]. The incorporation of Fe could modulate the adsorption energy of OH on the catalysts. As shown in Fig. S13, the high-resolution O 1s spectrum can be clearly defined into four characteristic peaks of oxygen atoms (O1–O4) at binding energies of 530.2, 531.5, 532.2 and 533.0 eV, corresponding to a typical O from O^2– (O1), O from the OH^– (O2), hydroxyl groups or surface-adsorbed oxygen (O3) and surface-adsorbed water molecules (O4), respectively [18, 33, 34]. Based on the analyses of FT-IR and XPS results, it can be concluded that the formed product should be Fe-Co oxyhydroxide.

The electrochemical properties of the as-synthesized catalysts were evaluated in a standard three-electrode system using 1.0 mol L^–1 KOH as the electrolyte. According to the linear sweep voltammetry (LSV) curves in Fig. 5a, 3D-FeCoOOH/CC exhibits the best catalytic activity for OER. The obtained 3D-FeCoOOH/CC shows an onset potential (1.4 V) smaller than those of 3D-CoOOH/CC and IrO₂. The potential required to reach a current density of 10 mA cm^–2 and the current density at an overpotential of 300 mV has been widely investigated to evaluate the catalytic activity for OER [38, 39]. It can be seen from Fig. 5b (black columns) that to drive the current density of 10 mA cm^–2, 3D-FeCoOOH/CC requires a minimum overpotential of 259 mV. As illustrated in Fig. 5b (red columns), 3D-FeCoOOH/C exhibits a current density of 137 mA cm^–2, which is much larger than those of 3D-CoOOH and IrO₂. It can be concluded that 3D-FeCoOOH/C exhibits an outstanding OER performance in comparison to the performance for others. Additionally, 3D-FeCoOOH/CC exhibits the highest mass electrocatalytic activity for OER (Fig. S14).

The OER reaction kinetics of FeCoOOH/CC was investigated by the Tafel slope according to the Tafel equation (η = blogj + a, where a, b, and j are a constant, the Tafel slope, and the current density, respectively) [25, 29]. As shown in Fig. 5c, the Tafel slope of 3D-FeCoOOH/CC is only approximately 34.9 mV dec^–1, which is lower than those of 3D-CoOOH/CC (59.2 mV dec^–1) and IrO₂ (80.0 mV dec^–1), indicating the most favorable OER kinetics [39]. To compare the catalytic activity of 3D- FeCoOOH/CC with those for the reported Co-based electrocatalysts and other state-of-the-art electrocatalysts, the overpotential at 10 mA cm^–2 and the corresponding Tafel slope are listed in Table S3. It is clearly observed that the as-synthesized 3D-FeCoOOH/CC exhibits superior catalytic activity compared to those of the reported electrocatalysts.

In addition to the catalytic activity, stability is of importance as it is required for commercial applications [2]. The multi-current step chronopotentiometric curve of 3D-FeCoOOH/CC in Fig. 5d shows that the potential rapidly stabilizes at 1.53 V and remains steady for 20 min. An analogous response is observed through the consequent chronopotentiometric processes, which reflects the superior mass transportation property of the 3D-FeCoOOH/CC electrode [20, 40]. The long-term stability was further evaluated by chronopotentiometric method at 20 mA cm^–2, and the corresponding curves are shown in Fig. 5e. It is determined that the overpotential of 3D-FeCoOOH/CC shows almost no change after 100 h of testing, while IrO₂ shows noticeable degradations only after 20 h of testing, indicating the superior stability of 3D-FeCoOOH/CC compared to that of IrO₂. After 100 h of testing, the SEM images in Fig. S15 show that the 3D-FeCoOOH/CC retains its porous structure without noticeable changes. This indicates that 3D-FeCoOOH/CC is structurally robust and highly stable, which is favorable for practical application.

To investigate the reasons for the high catalytic activity of 3D-FeCoOOH/CC, the electrochemically active surface area (ECSA) was measured by the double-layer capacitance (C_dl) method in 1.0 mol L^–1 KOH solution in no-Faradaic current potential range (1.18–1.26 V) [21]. As shown in Figs. S16a and S16b, the current density of 3D-FeCoOOH/CC at 1.22 V is nearly five times that of 3D-CoOOH/CC. The different assembly of these nanoparticles could be the key to this enhancement. They are a part of a thin conformal coating in 3D-CoOOH/CC in contrast to the micro go shape in 3D-CoOOH/CC, where the center of the go pieces might be not accessible. The C_dl values of 3D-FeCoOOH/CC and 3D-CoOOH/CC shown in Fig. S16c are 88.5 and 16.3 mF cm^–2, and the corresponding ECSA values are 2212.5 and 407.5 cm², respectively, indicating more exposed active sites. The higher ECSA of 3D-FeCoOOH/CC further confirms the compositional and structural advantages in terms of the exposure of catalytic active sites [41].

The electrochemical impedance spectroscopy (EIS) and the corresponding circuit model fitting analysis were further performed (Fig. S16d) at an overpotential of 300 mV. The equivalent circuit of EIS is comprised of a solution resistance (R_s), charge-transfer resistance (R_ct), and constant phase element (CPE) [23]. The R_ct of 3D-FeCoOOH/CC is 0.797 Ω, which is lower than that of 3D-CoOOH/CC (5.636 Ω), indicating a faster electron transfer in 3D-FeCoOOH/CC. The high conductivity of 3D-FeCoOOH/CC plays a key role in enhancing its OER performance, particularly at low overpotentials [11].

The above experimental results demonstrate that 3D-FeCoOOH/CC exhibits excellent electrocatalytic performances for OER. This could be attributed to the following features. First, the increased ECSA, conductivity and mass transportation caused by the incorporation of Fe led to its enhanced OER performance. Second, the strong interaction between Co and doped Fe modulated the electron density and the adsorption energies of OH on the catalysts, leading to the enhanced OER performance. Third, the unique 3D porous structure of 3D-FeCoOOH/CC possessed large surface area and it favored the exposure of more active sites, increase in the contact area of electrolyte and catalyst, and release of generated oxygen. Fourth, the unique 3D porous structure was highly robust and stable, leading to a good long-term stability. Owing to these unique features, the as-synthesized 3D-FeCoOOH/CC electrode exhibited superior electrocatalytic performances for OER.

In summary, we have successfully developed an efficient strategy for the assembly of 3D-FeCoOOH/CC. The 3D-FeCoOOH/CC possessed a low overpotential of 259 mV at 10 mA cm^–2 and small Tafel slope of 34.9 mV dec^–1 for OER, thereby outperforming Ir- and Ru-based catalysts. The as-synthesized 3D-FeCoOOH/CC also exhibited excellent stabilities after continuous electrolysis in alkaline electrolyte for more than 100 h with negligible activity decay. The 3D-FeCoOOH/CC with excellent OER catalytic activity has excellent potential for application in water splitting. Furthermore, the synthetic strategy is also of significance for the fabrication of a series of Co-based electrode materials with the dopant of other earth-abundant transition elements.

This work was supported by the National Natural Science Foundation of China (21475071), the Taishan Scholar Program of Shandong (ts201511027), and the Natural Science Foundation of Shandong Province (2018GGX102030). Han-Pu Liang is thankful for support from the "Hundred Talent Program" of Chinese academy of Sciences (CAS) (RENZI[2015] 70HAO, Y5100619AM), DICP and QIBEBT (UN201804), Dalian National Laboratory For Clean Energy (DNL), CAS, and Research Innovation Fund (QIBEBT SZ201801).