1 Introduction

Actually, global demand for energy is increasing rapidly because of population and economic growth particularly in emerging countries. The whole human population over the world depends generally on the energy produced from petroleum derivatives. However, the amount of carbon-based fuels diminishes and pollution of environment have drawn the attention of researchers toward sustainable and nonconventional sources and fuels [1, 2]. To resolve the energy crisis and reduce the emission of pollution, the utilization of hydrogen (H2) is an essential renewable energy carrier. In the fuel cell, it is combined to oxygen for generating electricity and forms only water, which makes it environmental friendly [3]. Production of H2 from steam reforming of methane is a dominant method and produces H2 in large quantity altogether with gases such as carbon dioxide and carbon monoxide [4]. In the same context, water electrolysis is one of cleanest ways available capable of providing large amounts of the renewable hydrogen. In practice, the efficiency of water electrolysis is limited by kinetically slow oxygen evolution reaction (OER), since the anodic reaction is usually slower than cathodic reaction [5,6,7]. The large anodic overpotential during the electrolysis and high activation in aqueous solution are the main factors that affect the OER rate. Therefore, the main objective of the current research is to understand the OER kinetics and reduce the over potential in alkaline media. Many works have been made in the past to use electrocatalysts for enhancing the electrochemically active area. Traditionally, various mixed oxides such as IrO2, RuO2, and TiO2 have been used frequently as catalysts for OER, which often showed the best catalytic activity and less stability under alkaline electrolyte as well as high cost [8]. During recent years, increasing attention has been directed now to search high active and stable OER catalysts based on cheaper metals such as oxides and hydroxides based of transition metals (Ni, Co, Fe, etc.) [9,10,11].

Perovskite-type oxides with the general formula ABO3 are estimated as active electrocatalysts for OER reaction, owing to their high catalytic activity [12]. The partial replacement of A or/and B sites by suitable metal ions is responsible for the thermal resistance and catalytic activity, respectively [13]. Several studies have been carried out on the OER performance of pure and substituted of LaFeO3 perovskites. Sankannavar et al. have reported that substitution of lanthanum by calcium increases the average oxidation state of iron from Fe3+ to Fe4+ and creates oxygen vacancies [1]. Another works have mentioned that Fe4+ ion is the OER active site on iron oxides [14,15,16]. Moreover, Suntivich et al. [17] have reported that the high OER activity of perovskite oxides is due to eg occupancy close to unity of B site transition metal. Zinc was used as dopant in several works. This choice is justified by its abundance, ability to produce extrinsic defects [18] and its important role in enhancing electrocatalytic performances for OER [19].

Despite all these advantages, LaFe1-xZnxO3 oxides have been subject of few works. Huang et al. have studied the sensitivity of LaFe1-xZnxO3 to formaldehyde, shown that Zn-doping enhances the gas sensitivity of LaFeO3 [20]. Two years later, Dong et al. have investigated the photocatalytic properties of these materials for the degradation of methylene blue. It has been found that the LaFe0.7Zn0.3O3 sample exhibits the highest degradation rate of 75% under irradiation time of 150 min [21]. Thereafter, Bhat et al. have synthesized using sol–gel auto-combustion process, characterized magnetic and dielectric properties of LaFe1-xZnxO3 oxides [22]. To our best knowledge, there is no report in the literature on the electrocatalytic properties of LaFe1-xZnxO3 for the OER. In our present work, first row transition metal (Fe and Zn) based perovskite materials LaFe1-xZnxO3 (0.0 ≤ x ≤ 0.3) were synthesized by a typical sol–gel route. The electrocatalytic activity and the kinetics of the OER on the doped zinc lanthanum ferrites electrodes were investigated using linear sweep voltammetry (LSV), cyclic voltammetry (CV), Tafel slopes, and electrochemical impedance spectroscopy (EIS).The electrochemical stability analysis of LaFe1-xZnxO3 electrodes was also performed under OER conditions over 1000 cycles.

2 Materials and methods

2.1 Preparation of LaFe1-xZnxO3 catalysts

LaFe1-xZnxO3 powders, which can be named LF (x = 0.0), LFZ10 (x = 0.1), LFZ20 (x = 0.2), LFZ30 (x = 0.3), were prepared by citrate sol–gel method with employed lanthanum, zinc, iron nitrates as starting materials and citric acid as a complexant agent. Precursors La(NO3)3·6H2O, Fe(NO3)3.9H2O, Zn(NO3)2.6H2O dissolved in ethanol were well-mixed under vigorous stirring, subsequently, citric acid was added into the above mixture. The resulting homogeneous mixture was heated at 60–70 °C with continuous stirring until a violet red transparent solution was obtained, becomes gradually viscous, and later the amorphous gel was formed. To remove residual ethanol, the gel was dried in an oven at 100 °C for 24 h. Finally, the LaFe1-xZnxO3 powders were further calcined at 750 °C under air atmosphere for 6 h.

2.2 Catalysts characterization

The crystallographic information on the LaFe1-xZnxO3 (0.0 ≤ x ≤ 0.3) oxides were recorded by a powder X-ray diffractometer (D8 ADVANCED-BRUCKER) equipped with CuKα as the radiation (λ = 1.542 Å) at a sweep rate of 2° min−1 from 10 to 90°. X’pert High-Score plus software (version 2.1.0) was used for the phase identification of powders by comparing with the International Centre for Diffraction data. Moreover, morphological aspects of the powders were examined by scanning electron microscopy (SEM) (model VEGA 3 TESCAN).

2.3 Electrode preparation

The oxide electrocatalysts prepared in powder form were reproduced in the form of film on a pretreated nickel support by an oxide-slurry painting method. Prior to use, Ni plates (2 × 1 cm) were etched for 10 min in concentrated HCl, washed with double distilled water, degreased in acetone, and then dried in air. The powder was dispersed in ethanol by subsequent ultra-sonication of the mixture for 30 min and then stirred in an oven at 90 °C. A pure nickel plate was coated by a thin layer of oxide powder, then heating at 100 °C for 2 h in an electrical furnace to sinter the binder particles. The oxide load on the electrode was ̴15 mg cm−2.

2.4 Electrochemical measurements

A three-electrode single compartment pyrex glass cell was used to carry out electrochemical investigations with a potentiostat–galvanostat (model Parstat 4000). The reference and auxiliary electrodes are Hg/HgO and Pt, respectively. All potential values mentioned in the text have been referred to this reference electrode. Procedure followed in the study of LSV, CV, anodic Tafel polarization has already been described elsewhere [23, 24]. The EIS study of the oxide film electrodes in 1 M KOH has been carried out with an ac voltage amplitude of 5 mV over the frequency range of 0.01–104 Hz at varying potentials (0.60–0.68 V). The stability of LF and LFZ10 electrodes was studied after 1000 cycles of continuous CV scans with a scan rate of 20 mV/s.

3 Results and discussion

3.1 Structural characteristics of LaFe1-xZnxO3 catalysts

Figure 1 presents XRD patterns of the calcined samples LaFe1-xZnxO3 (0 ≤ x ≤ 0.3) at 750 °C for 6 h. The samples LF and LFZ10 exhibit a pure perovskite phase with a cubic structure (JCPDS:01-075-0541). At high zinc content (x ≥ 0.2), new peaks appear with weak intensity corresponding to ZnFe2O4 (JCPDS: 01-073-1963), ZnO (JCPDS: 01-075-1526), and La2O3 (JCPDS: 00-002-0688). This result is consistent with a previous report which indicates that LaFe1-xZnxO3 exhibit a pure cubic perovskite up x = 0.1 [21]. A slight shift of peaks can be observed with increasing the amount of zinc, which is probably due the difference between ions radius of Fe3+ (0.64 Å) and Zn2+ (0.74 Å).

Fig. 1
figure 1

X-ray diffraction patterns of synthesized LaFe1-xZnxO3 (x = 0.0–0.3) after calcination at 750 °C, c: cubic phase; *: ZnFe2O4; −: ZnO; X: La2O3

Full size image

3.2 Scanning electron microscopy

The morphologies of the pure and doped of LaFeO3 powders studied by SEM are shown in Fig. 2. The micrographs images indicate that particles shapes are not well defined. It seems to consist of macro-agglomerations of numerous nano-crystallite particles with varying shape and grain sizes (particle–particle interactions). In this case, all the compounds show a nonuniform distribution with some clustered particles. The formation of agglomerates is likely due to the nature of the solvent used in the preparation of the gel. Moreover, the surface morphologies of these zinc-doped ceramics are dense, which is probably due to the particles that have a high tendency of agglomeration with interconnected structure.

Fig. 2
figure 2

SEM images of LF: a LFZ10; b LFZ20; c powders calcined at 750 °C

Full size image

3.3 Electrocatalytic properties of LaFe1-xZnxO3 catalysts

3.3.1 Oxygen evolution activity

The linear sweep voltammograms of LaFe1-xZnxO3 (0.0 ≤ x ≤ 0.3) electrodes in the potential region 0.0–0.7 V vs. Hg/HgO in 1 M KOH with a scan rate of 50 mV s−1 are presented in Fig. 3. It is observed that these electrodes have a qualitatively similar behavior. Each voltammogram exhibits only one anodic peak (520 mV ≤ Epa ≤ 576 mV) related to the Ni(II)/Ni(III) redox couple [25, 26]. The shift in the position of redox peaks at lower potential region may be ascribed to the synergistic interaction between Ni metal and LaFe1-xZnxO3 catalysts [25,26,27]. Thus, pure Ni used as the support in the present study may produce oxidation-reduction (Ni2+/Ni3+) reaction during the anodic cycling condition owing to the electrolyte contact through pores, cracks, and so on [27, 28]. However, the observed redox peaks on the oxide films on Ni have not been originated from the catalytic films but were produced from the oxidation of the Ni support [28]. In all cases during a forward scan, these peaks appear just prior to the onset of oxygen evolution with the slight shift toward high potentials. Both the OER onset potential and the overpotential (defined as the potential at the current density of 10 mA cm−2) increase according to the following order: LFZ10 < LFZ20 < LFZ30 < LF, which pointed out clearly that the better electroactivity toward the OER can be achieved with LFZ10 electrode. The value of overpotential required to achieve a current density of 10 mA cm−2 for LFZ10 is ~677 mV which is comparable to data already reported in the literature for analogous systems, La0.2Sr0.8FeO3 (~0.66 V) [29], CaFeO3 (~0.81 V) [1], and GdFeO3 (~0.68 V) [30].

Fig. 3
figure 3

Linear sweep voltammetry measurements of LaFe1-xZnxO3 powders in 1 M KOH, at a scan rate of 50 mV/s; inset shows the OER current at 0.66 V vs. Hg/HgO

Full size image

On the other hand, the addition of zinc allowed to raise the current density which is approximately three times higher for LFZ10 than LF at E = 0.66 V. Indeed at this potential, the LFZ10 catalyst delivers a highest current density ~9.02 mA cm−2 compared with all compositions. This is possibly due to the substitution of a transition metal iron by zinc leading to the formation of cation vacancies, which are the result of the high solubility of Zn in KOH solution. This result was also reported by previous works indicating that the addition of zinc generates metal vacancies, defective sites, and hydroxylation, which enhances the OER performance of catalysts [31, 32].

At higher zinc content (x > 0.1), it decreases strongly to reach a current density ~3.0 mA cm−2 for x = 0.3, which is close to that obtained for LF. This indicates probably that the apparition of the secondary phase ZnO which becomes more intense for higher zinc content has an inhibition effect on OER performance [33].

Recently, similar result was also reported in La1-xSrxNiO3 [34]. It has been shown that the SrCO3 impurity is inactive for the OER and blocks the active site of electrocatalysts, which negatively impacts the electrocatalytic activity.

Kinetic parameters of the OER, for pure and zinc-doped perovskite electrodes in 1 M KOH at 25 °C, are recorded at a slow scan rate 0.2 mV s−1 (Fig. 4). The obtained Tafel plots of LF, LFZ10, and LFZ20 electrocatalysts are 89, 52, and 64 mV dec−1, respectively. This result indicates that the electrode LFZ10 exhibits a better performance in the OER and is consistent with LSV findings.

Fig. 4
figure 4

Tafel plots of LaFe1-xZnxO3 powders

Full size image

3.3.2 Electrochemical impedance spectroscopy

EIS was used to study the OER kinetics properties exhibited by the oxide films. A typical complex plane diagram for the active oxide LaFe1-xZnxO3 (0.0 ≤ x ≤ 0.2) at a constant potential E = 0.64 V is presented in Fig. 5. The experiment data are represented by discrete symbols and the simulated impedance reponse is represented by continuous line. The two circuit description codes, Rs(QfRf)(QdlRct) and LRs(QfRf(CdlRct)) for LF, LFZ10, and LFZ20, respectively, are used to simulate the EIS data and estimate the corresponding circuit parameters. A good agreement is obtained between simulated and experimental curves.

Fig. 5
figure 5

Complex impedance (Nyquist) of LF, LFZ10, and LFZ20 electrodes in 1 M KOH solution at 25 °C, E = 0.64 V. The experiment data are represented by discrete symbols and the simulated impedance reponse is represented by continuous line

Full size image

The best fit values of the circuit parameters are shown in Table 1. It can be observed that the resistance of charge transfert increases with the substitution degree from x = 0.1 to 0.2, which is in agreement with the findings from LSV indicating that the higher electroactivity is achieved for LFZ10.

Table 1 Values of the equivalent circuit parameters for oxide electrodes in 1 M KOH at 0.64 V
Full size table

3.3.3 Stability test

Long-term stability is an important feature for catalysts that can be used in practical applications. In this aim, LF and LFZ10 electrodes stability was studied under OER conditions. Figure 6 presents the overpotential of LF and LFZ10 required to achieve 3 and 10 mA cm−2, respectively, during 1000 continuous cycles. The choice of using two different current densities is justified by the fact that it is not possible to estimate the overpotential at the common current density for the two samples in the first cycle. In the case of LF sample, the overpotential decreases continuously during the 1000 cycle stability test, which is probably due to the porous nature of this electrode [35]. Two distinct regions are observed corresponding to 100–400, 400–1000 cycles, with a decrease of the overpotential by 60 and 110 mV, respectively. This indicates that after 1000 cycles, this electrode was activated, remains stable, and even more it performs better in the OER domain. For the second sample LFZ10, after the first 100 consecutive cycles, an increase of overpotential of 19 mV is required to achieve a current density of 10 mA/cm2, indicating that in this interval the electrode has been desactivated. This is probably due to the surface coverage of oxygen bubbles, formation of film cracks, and so on [36].

Fig. 6
figure 6

Change of cell voltage at various points during the 1000 cycle stability tests for LF at 3 mA cm−2 and LFZ10 at 10 mA cm−2

Full size image

This overpotential decreases slightly after the cycle 200, from 19 to 11 mV corresponding to the 1000th cycle, indicating that this electrode has been activated. For comparison an overpotential by 3.41 mV is required for La0.2Sr0.8FeO3 electrode [29], to achieve 3 mA cm−2 after 20 cycles, which shows clearly that the very small change in overpotential between 11 and 19 mV, required for LFZ10 to achieve a current density of 10 mA/cm2 over a large number of cycles indicates a relatively excellent stability of LFZ10 electrode.

4 Conclusion

In summary, Zn-doped LaFeO3 were successfully synthesized by the sol–gel method, employing nitrate salts of lanthanum, iron, zinc as cations precursors, citric acid as a chelating agent. According to X-ray powder diffraction, LaFe1-xZnxO3 oxides exhibit a single phase at 750 °C in the range (0.0 ≤ x < 0.2). Two detectable secondary phases ZnFe2O4 and ZnO were observed for x = 0.2 indicating that a solubility is reached for x < 0.2. The morphology of samples examined by SEM show a nonuniform grain size distribution and very well pronounced agglomeration of powders. The electrochemical investigation revealed that the substitution of Fe by Zn in the base oxide increased the electrocatalytic activity toward OER. The value being highest with 0.1 mol Zn substitution at E = 0.66 V. Moreover, it exhibits a relatively excellent stability after 1000 continuous cycles. Therefore, the LFZ10 oxide can be proposed as an electrode material with a good performance for OER of alkaline water electrolysis process.