Introduction

Electrochemical nitrate reduction reaction (NO3RR) offers a promising solution to reduce nitrate pollution and produce high-value-added ammonia (NH3) [1,2,3]. However, the NO3RR is kinetically sluggish, with the overpotential being usually exceeding 1 V [4,5,6]. Furthermore, the large driving force of the anodic oxygen evolution reaction (OER) prevents the NO3RR from being applied on a large scale [7,8,9]. Therefore, tremendous efforts have been made to develop highly efficient NO3RR and OER electrocatalysts [10, 11]. Non-noble transition metal catalysts are the ideal choice owing to their low cost, high efficiency, and easy modification [12,13,14,15,16].

The charge transfer steps are usually the rate determining step in the NO3RR and OER [17,18,19,20]. However, the mass transfer steps are more likely to control the reaction rate at the high overpotential [21, 22]. Therefore, understanding the influence of the mass transfer in the NO3RR and OER is important for further large-scale industrial applications. The nitrate and ammonium ion transfer primarily involves the liquid and solid phases in the NO3RR. Furthermore, the oxygen transfer requires gas, liquid, and solid phases. One-dimensional (1D) nanomaterials exhibit outstanding mass transfer properties owing to the plenty of open spaces among the adjacent 1D nanostructure. He et al. [23] found that the copper/cobalt array could efficiently convert NO3− to NH3. Furthermore, the current density was large because the mass transfer was excellent. Liu et al. [24] demonstrated that the 1D carbon nanotube can enhance the OER performance of the CoMn-layered double hydroxides. This may be because of the easier diffusion of oxygen from the surface of the as-prepared nanomaterials. Thus, achieving the high NO3RR and OER activity by designing 1D nanomaterials with different microstructures is promising. Furthermore, it is essential to determine how the 1D microstructure influences interfacial mass transfer. The electrospinning could easily change the 1D structure by adjusting the related parameters, such as voltage, distance, and viscosity [25,26,27].

In this research, we systematically investigated the influence of the gas–liquid–solid triphase interfacial mass transfer on the NO3RR and the OER using a series of nickel and iron-loaded nanofibers (NiFe-NFs) materials, which were synthesized via electrospinning and pyrolysis methods. The diameter of the as-prepared 1D NFs could be controllably adjusted by varying the preoxidative heating rate. Among these nanomaterials, the NiFe-NFs-2 (under pre-oxidative heating rate of 2 °C/min) displayed a surprising NO3RR activity, a Faradaic efficiency of 94.3%, and a large NH3 yield of 10,489 µg/(h·mg), showing an excellent OER performance with an overpotential of 307 mV at the current density of 10 mA/cm2 and a small Tafel slope of 44.3 mV/dec. Experiments and theoretical calculations indicated that the diameter of the 1D NFs could change the reactant transfer, thereby improving the electrochemical performance.

Results and Discussion

The NiFe-NFs were synthesized via electrospinning and two-step heat treatment (Fig. 1a). First, the precursor NFs were prepared via electrospinning and were then preoxidized at 250 °C in air atmosphere and carbonized at 800 °C in an Ar/H2 atmosphere to obtain NiFe-NFs. The heating rate during the preoxidation could control the diameter of the NiFe-NFs. Thus, the different heating rates of 1 °C/min, 2 °C/min, 3 °C/min, 4 °C/min and 5 °C/min were applied during the synthesis process, and the as-prepared NiFe-NFs were named NiFe-NFs-X (X = 1, 2, 3, 4, 5; X represents the heating rate of preoxidation).

Fig. 1
figure 1

a Schematic of the NiFe-NFs-X synthesis process. b SEM image of NiFe-NFs-2, the inset shows the diameter distribution of nanofibers. c TEM image of NiFe-NFs-2 in the inset shows the diameter distribution of nanoparticles. d HRTEM image of NiFe-NFs-2. e XRD patterns, f Raman spectrum, and g TG curves of as-prepared samples

Full size image

Scanning electron microscopy (SEM) was used to analyze NiFe-NFs-X (Figs. S1, S2) to confirm the effect of the preoxidation rate on the catalyst morphology. The precursor NFs displayed a smooth surface with a uniform diameter of ~ 300 nm (Figs. S1, S2). After the preoxidation, the gas released during the precursor nanofibers crystallizes and decomposes, and the diameter of the NFs was considerably reduced with the different preoxidation heating rates (Fig. S2). With the decrease in the preoxidation rate, the diameter of NFs gradually decreases. The preoxidation heating rate of 2 °C/min makes the diameter of the NFs smaller. The carbonization further reduced the diameter of the NFs and exposed the NiFe nanoparticles to varying degrees. The SEM images revealed that the NiFe-NFs-2 has the smallest diameter of NFs with an average diameter of 45.3 nm (Figs. 1b, S3). Transmission electron microscopy (TEM) images (Figs. S4, 1c) revealed that the average nanoparticle size of NiFe-NFs-2 is ~ 24.65 nm, which is considerably smaller than those of other catalysts. Notably, the combination of NFs and nanoparticles in NiFe-NFs-X creates a rough surface structure that considerably increases the specific surface area of the catalyst and facilitates the mass transfer process of the reaction. Among them, NiFe-NFs-2 has the roughest surface structure, and it is speculated that it possesses better catalytic activity.

Furthermore, the high-resolution transmission electron microscopy (HRTEM) image revealed the lattice fringes at 0.213 nm and 0.185 nm assigned to the facets (111) and (002) of NiFe alloy, respectively. The X-ray diffraction (XRD) pattern of NiFe-NFs further confirmed the presence of NiFe nanoparticles on the nanomaterials. Prominent diffraction peaks 2θ = 43.76° and 50.92° matched the standard NiFe peaks (PDF No.98-063-2933) (Fig. 1e). Moreover, the C peak at 25° gradually decreased with decreasing rate of NiFe-NFs-X preoxidation. This is primarily due to the large number of NiFe nanoparticles covering the catalyst surface, which weakened the intensity of the C peak signal. The Raman spectrum of NiFe-NFs-X revealed two peaks at 1300 and 1580 cm−1 corresponding to the D and G bands of the carbon materials. The intensity ratios ID/IG of NiFe-NFs-X were similar (Fig. 1f), implying that the graphitization degrees of the samples were similar. The preoxidation heating of precursor NFs was studied using the thermogravimetric (TG) method. The calcined TG curves depicted the fiber heating to 250 °C in air atmosphere at varying heating rates ranging from 1 to 5 °C/min (Fig. 1g). TG curves revealed that the weight loss rate was inversely proportional to the preoxidation heating rate, implying that the fiber weight loss rate was smaller at a larger heating rate. This is primarily because the preoxidation process increases the amount of C–O and C=O on the catalyst surface, and the slower the heating rate, the higher the preoxidation degree and, thus, the greater the weight loss rate. Therefore, the surface roughness increases with decreasing heating rate from 2 to 5 °C. However, with a further decrease in heating rate, the particles will be considerably exposed, and the roughness will decrease due to agglomeration at high temperatures.

To further determine the chemical composition and elemental states of NiFe-NFs, X-ray photoelectron spectroscopy (XPS) was performed. The characteristic signatures for C, O, N, Ni, and Fe element were visible in the survey scan spectrum of NiFe-NFs-X (Fig. S5). Table S1 presents the atomic ratio of NiFe-NFs-X. Compared to other catalysts, NiFe-NFs-2 has the highest Ni and Fe content (Ni: 16.64% and Fe: 11.85%) and the lowest C content (C: 42.92%), indicating that the NiFe nanoparticles are fully exposed to the C surface. The C 1s spectrum of NiFe-NFs-X reveals four fitted peaks located at 284.8, 285.6, 286.9, and 289.2 eV corresponding to C=C, C–N, C=O, and C–O, respectively (Figs. 2a, S6a–d) [28,29,30]. The decomposition of the N 1 s spectrum of NiFe-NFs-X is fitted with three peaks at 398.3, 399.3, and 400.8 eV corresponding to pyridinic N, pyrrolic N, and graphitic N, respectively (Figs. 2b, S7a–d) [31, 32]. The Ni 2p spectrum of NiFe-NFs-X deconvoluted into Ni0 (852.9 and 870.1 eV) and Ni2+ (855.6 and 873.6 eV) (Figs. 2c, S8a–d) [33,34,35]. The Fe 2p spectrum of NiFe-NFs-X deconvoluted into Fe0 (707.4 and 720.5 eV), Fe2+ (710.3 and 723.8 eV), and Fe3+ (712.5 and 725.1 eV) (Figs. 2d, S9a–d) [36,37,38]. The XPS results revealed that the C 1s and N 1s regions of NiFe-NFs-X are similar, indicating the consistent in properties of carbon NFs [28]. When comparing NiFe-NFs-2 with NiFe-NFs-X, the peaks of Ni0 and Fe0 were more prominent. This is primarily due to the NFs of NiFe-NFs-2 and the exposed NiFe nanoparticles are more easily reduced by Ar/H2 gas.

Fig. 2
figure 2

High-resolution XPS spectra of a C 1s, b N 1s, c Ni 2p, and d Fe 2p of NiFe-NFs-2

Full size image

The NO3RR performance of the NiFe-NFs-X catalysts was investigated using an H-type cell with a three-electrode system in 0.05 mol/L KOH, 0.1 mol/L K2SO4, and 0.5 mol/L KNO3 mixed electrolyte. The linear sweep voltammetry (LSV) curves of NiFe-NF-X were plotted to compare the catalyst activity (Fig. 3a). The results revealed that NiFe-NFs-2 has a lower onset potential than other catalysts. At the same potentials, the NiFe-NFs-2 exhibited a higher current density than other catalysts.

Fig. 3
figure 3

a LSV curves of NiFe-NFs-X in 0.05 mol/L KOH, 0.1 mol/L K2SO4 and 0.5 mol/L KNO3 mixed electrolyte. b FE and yield rate of NH3 for the prepared NiFe-NFs-X at − 0.654 V (vs. RHE). c Cycling tests of NiFe-NFs-2 at − 0.645 V. d OER polarization curves of various catalysts in 1.0 mol/L KOH with IR compensation. e Overpotential of different catalysts at 10 and 50 mA/cm2 for OER. f LSV plots before and after chronoamperometric tests of OER

Full size image

The LSV curves of NiFe-NFs-2 in mixed electrolyte with or without 0.5 mol/L NO3− were investigated. The NiFe-NFs-2 exhibited a larger current density and lower onset potential in the electrolyte system containing NO3− than without NO3−, indicating the occurrence of NO3RR in the electrolytic system (Fig. S10). Subsequently, the Faradaic efficiency and yield of NH3 of NiFe-NFs-2 at various potentials were further investigated using chronoamperometry (CA), and the NH3 generated was quantified using the indophenol blue method (standard curve in Fig. S11). The current density and NH3 yield gradually increased with the increase of cathode potential (Fig. S12). However, the Faradaic efficiency of NH3 declined at higher potential due to the severe side reaction of hydrogen evolution. Simultaneously, bubbles can be observed on the working electrode during electrolysis. The results revealed that NiFe-NFs-2 had the best performance with a Faradaic efficiency of 94.3% and an ammonia production performance of 10,489 µg/(h·mg) at − 0.645 V vs. RHE. Meanwhile, NiFe-NFs-2 exhibited the highest Faradaic efficiency and catalytic activity than other catalysts (Figs. 2b, S13). To further validate the N source of NH3, we tested without NO3− and with 0.5 mol/L KNO3 electrolytes. Only trace amounts of NH3 were detected without NO3− electrolyte, and its current density was considerably smaller than that in the 0.5 mol/L KNO3 electrolyte (Fig. S14). The above results showed that the rough surface of NiFe-NF-2 favors mass transfer.

Durability is another important characteristic of electrocatalyst, and the endurance cycle test of NiFe-NFs-2 was carried out at − 0.645 V. After six cycles (0.5 h each) (Fig. 3c), indicating that NiFe-NFs-2 had good electrochemical stability. There were no significant fluctuations in ammonia yield and Faradaic efficiency. In the long-term stability test of 12 h, we can see a slight decrease in the current density of the reaction. (Fig. S15). SEM showed that the slight decrease in current was mainly due to the slight shedding of particles. (Fig. S16a). After the NO3RR stability tests, the XPS confirmed that the valence chemical states were well-preserved (Fig. S17). Considering the favorable NO3RR activity and good stability, NiFe-NFs-2 can be a promising catalyst for efficient NO3RR electrocatalysis.

Furthermore, the OER electrocatalytic activity of the prepared catalysts was evaluated using LSV at a scan rate of 10 mV/s in 1.0 mol/L KOH electrolyte. The LSV curves (Fig. 3d) showed that NiFe-NFs-2 had the lowest onset potential. Compared to other materials, NiFe-NFs-2 exhibited a low overpotential of 307 mV at 10 mA/cm2, which was lower than those of NiFe-NFs-1 (336 mV), NiFe-NFs-3 (352 mV), NiFe-NFs-4 (363 mV), and NiFe-NFs-5 (392 mV). The OER overpotential of the 5 kinds of catalysts was compared at high current density (Fig. 3e). It was found that the OER activity of NiFe-NFs-2 remained outstanding at a high current density of 50 mA/cm2. The NiFe-NFs-2 showed strong electrochemical durability as verified by the chronoamperometric response (Fig. S18). The favorable stability of NiFe-NFs-2 can result from its robust structure, as confirmed by the well-preserved morphology and the valence chemical states after the stability tests (Figs. S16b, S19). The LSV curve of the before and after stability test was not much different (Fig. 3f). This result suggested that the NiFe-NFs-2 also has better stability for OER: the specific structure and outstanding mass transfer attributed to the excellent performance.

To further investigate the excellent performance of NiFe-NFs-2, we found that the Tafel slope (Fig. 4a) of NiFe-NFs-2 was 44.2 (mV/dec), which was lower than those of NiFe-NFs-1 (56.2 mV/dec), NiFe-NFs-3 (58.1 mV/dec), NiFe-NFs-4 (66.7 mV/dec), and NiFe-NFs-5 (78.8 mV/dec), indicating that NiFe-NFs-2 had the fastest OER reaction kinetics. Furthermore, the Tafel slope value was close to 40 mV/dec, indicating the rate determining step was M + OH− − e− = M–OH. Thus, OH− mass transfer would improve the OER performance. Furthermore, electrochemical impedance spectroscopy tests were performed, where the intersection of the real axis with the starting point of the Nyquist plots represents the solution resistance (R1), and the semicircle diameter represents the charge transfer resistance (R2). The Nyquist plot showed that the internal resistance of NiFe-NFs-2 is as low as 12.1 Ω, considerably lower than those of NiFe-NFs-1 (29.8 Ω), NiFe-NFs-3 (33.4 Ω), NiFe-NFs-4 (50.2 Ω), and NiFe-NFs-5 (75.9 Ω) (Figs. 4b, S20). Consequently, NiFe-NFs-2 had a lower charge transfer resistance and a faster charge transfer rate. Theoretically, the electrochemically active surface area (ECSA) is linearly related to the Cdl value, so the ECSA of different materials can be obtained by comparing their Cdl values. The Cdl value of NiFe-NFs-2 (5.34 mF/cm2) was significantly greater than those of the other prepared catalysts (Figs. 4c, S21). The Brunauer–Emmett–Teller measurements showed that NiFe-NFs-2 has a surface area of 461.547 m2/g, which is higher than those of NiFe-NFs-1 (444.557 m2/g), NiFe-NFs-3 (427.793 m2/g), NiFe-NFs-4 (360.151 m2/g) and NiFe-NFs-5 (313.824 m2/g) (Fig. S22). The large specific surface area of the catalyst facilitates the escape of the gases generated by the catalytic reaction and increases the exchange rate of the electrolyte over the catalyst [39]. The suitable preoxidation heating rate provides a larger electroactive surface for the OER, resulting in excellent OER performance of NiFe-NFs-2. With the same load, the modulation of NF diameter exposes NiFe-NF-2 to more electrocatalytic active sites and a larger interfacial contact area with the electrolyte, thus showing a higher activity. Combined with experimental results and characterization results, we can conclude that due to the rough surface structure of NiFe-NPs-2, the catalyst has a large specific surface area, promoting the mass transfer process and producing a high catalytic activity.

Fig. 4
figure 4

a Tafel plots of different catalysts. b Nyquist plots of prepared samples at 1.55 V (vs. RHE). c Capacitive currents at different scan rates of different catalysts

Full size image

Conclusions

We successfully prepared NiFe nanoparticle-doped NFs by electrospinning technology and a two-step pyrolysis method. Morphological and structural characterization indicated that the NiFe-NFs-2 surface is rougher than the compared catalyst. The NiFe-NFs-2 excellent structure and properties of NiFe-NFs-2 contribute to a high FE (94.3%), ammonia yield (10,489 µg/(h·mg)) for NO3RR, and low onset potential (307 mV) for OER. Furthermore, Tafel and ECSA experiments confirmed that a rougher surface can benefit a highly exposed active site and rapid mass transfer process. This research has offered novel ideas by improving the mass transfer process to enhance the catalytic capacity in NO3RR and OER.