English

• Nitrate, a highly soluble and toxic pollutant, has been increasingly detected in various water sources, including agricultural runoff, industrial wastewater, and drinking water. This pollutant is directly linked to the adverse eutrophication phenomenon and other serious environmental issues that threaten both ecosystem balance and human health [1-3]. The development of highly effective denitrification strategies is desirable and receives considerable attention. Among these, the electrochemical nitrate reduction reaction (NO₃RR) emerges as a promising method to achieve this goal in a greener and more sustainable manner, due to the remarkable advantages of its renewable electricity-driven nature, application feasibility, and high efficiency [4-7]. Additionally, ammonia (NH₃), as the principal product during NO₃RR, has been widely proposed as the most potential candidate for a sustainable hydrogen energy carrier because of its high energy density (4.3 kW/kg) and economic affordability for storage and transport [8]. Therefore, the advancement of NO₃RR technology has gain increasing attention from both the environmental and energy communities.

  Electrocatalysts play a crucial role in the emerging technology of nitrate reduction, as they significantly influence the activity and product distribution of NO₃RR. A wide range of materials with unique structures and compositions has been explored to enhance the NH₃ yield rate and Faradaic efficiency (FE) of NO₃RR, including pure metals (e.g., Ru and Co) [9,10], metal alloys (e.g., CuPd) [11-13], metal oxides (e.g., Co₃O₄ and Fe₃O₄) [14,15], and carbon-based materials (e.g., single atomic catalysts) [16-18]. Despite considerable progress in developing materials for NH₃ production via NO₃RR, the critical relationship between the structure of catalysts and their performance in NO₃RR remains unclear. This lack of clarity is due to insufficient understanding of the reaction process, which significantly impedes the rational design and advancement of catalysts for efficient NH₃ production. It is generally recognized that NO₃RR can proceed via either a direct electron transfer mechanism (DET) [19,20] or an indirect hydrogen atom (^•H) transfer (HAT) mechanism [21,22], with each expected to lead to different NO₃RR performance. Previous studies have shown that a balance between the consumption and supply of ^•H generated from the Volmer step during the hydrogen evolution reaction (HER) within the HAT mechanism is critical for superior NH₃ yield via NO₃RR [23]. An excessive supply of ^•H may lead to a preferred occurrence of competitive HER, thereby reducing the FE of NH₃. Theoretically, the DET mechanism does not involve hydrogen atoms as key intermediates for the parasitic HER, potentially leading to higher NO₃RR efficiency due to reduced ^•H generation. The dominance of a particular mechanism is highly dependent on the catalyst property. Therefore, a comprehensive understanding of the intricate relationship between catalyst structure, reaction mechanism, and NO₃RR performance is of great importance.

  Herein, we designed and fabricated three types of free-standing electrocatalysts, building upon our previous work with minor modifications [24]. These catalysts included pure metallic Fe nanoparticles in-situ grown on iron foil (denoted as Fe/FF), pure Fe₃O₄ nanosheets on carbon cloth (Fe₃O₄/CC), and a mixture of metallic Fe nanoparticles and Fe₃O₄ nanosheets in-situ constructed on iron foil (Fe@Fe₃O₄/FF). By carefully examining and comparing their NO₃RR performance in terms of NH₃ yield, NH₃ selectivity, and FE, we employed a combination of electrochemical characterizations, quenching experiments, electron spin resonance (ESR) measurements, and theoretical calculations to discern the differences in catalyst structure and reaction mechanisms. Through this comprehensive analysis, we were able to elucidate the structure-sensitive mechanisms of DET versus HAT and their impact on NO₃RR performance. This investigation provides valuable insights into catalyst design for the electrochemical treatment of nitrate-laden wastewater, offering an important perspective on optimizing the efficiency and selectivity of the NO₃RR process.

  The experimental section is elaborated in Supporting information, which includes chemicals and reagents, preparation of catalysts, experimental setup, physicochemical and electrochemical characterizations, analysis methods, and density functional theory (DFT) calculations. The electrodes used in this study were fabricated using a similar method as described previously [24], which involved initial annealing in a static air atmosphere followed by electrochemical activation. Briefly, Fe@Fe₃O₄/FF was prepared by annealing Fe foil at 500 ℃ in a static air atmosphere, followed by electrochemical cathode activation at 10 mA/cm². Meanwhile, Fe/FF was produced using the same procedure but with the annealing temperature adjusted to 350 ℃. A two-step protocol was employed for synthesizing Fe₃O₄/CC, which consisted of electrodeposition of Fe(OH)₃ on carbon cloth followed by air annealing at 350 ℃. Physicochemical characterization techniques including scanning electron microscopy (SEM), transmission electron microscopy (TEM), and diffractometry (XRD) were utilized to identify differences in composition and structure among the Fe/FF, Fe₃O₄/CC, and Fe@Fe₃O₄/FF cathodes. Fig. 1a and Fig. S1 (Supporting information) illustrate that the surface of Fe@Fe₃O₄/FF appeared rougher in comparison to the smoother surface of Fe₃O₄/FF prior to electrochemical treatment, with a high density of nanoparticles dispersed on the original Fe₃O₄ nanosheets. This observation aligns with our previous findings that electrochemical reduction treatment facilitates the formation of Co nanoparticles from Co₃O₄ nanosheets on cobalt foil (CF) under the influence of an applied cathode potential [24]. Fig. 1b shows that Fe₃O₄/CC exhibited flat nanosheet morphology, similar to Fe₃O₄/FF before electrochemical treatment. A lower calcination temperature of 350 ℃ during Fe/FF synthesis resulted in fewer Fe₃O₄ nanosheets forming on FF (Fig. S2 in Supporting information), in contrast to the abundant and dense formation of Fe₃O₄ nanosheets on Fe₃O₄/FF (calcination at 500 ℃), which were then entirely reduced to metallic Fe without any remaining Fe₃O₄ through the same electrochemical treatment, as depicted in Fig. 1c. Moreover, TEM analyses (Figs. 1d and e) further supported these findings, revealing lattice fringes with interplanar spacings of 0.25 nm and 0.215 nm, corresponding to the (311) plane of Fe₃O₄ and the (110) plane of Fe, respectively, indicating the coexistence of metallic Fe and Fe₃O₄ in Fe@Fe₃O₄/FF. However, for Fe₃O₄/CC, only Fe₃O₄ lattice fringes were observed (Fig. S3 in Supporting information). The composition of the three electrocatalysts was also confirmed through XRD analysis (Fig. 1f), where crystalline phases of Fe and Fe₃O₄ were identified for Fe/FF and Fe₃O₄/CC, respectively. In contrast, Fe@Fe₃O₄/FF exhibited crystalline phases of both Fe and Fe₃O₄, suggesting that Fe@Fe₃O₄/FF comprises both metallic Fe and Fe₃O₄ nanosheets.

  Figure 1

  

  Figure 1.  SEM images of (a) Fe@Fe₃O₄/FF, (b) Fe₃O₄/CC, and (c) Fe/FF. (d, e) HRTEM images of Fe@Fe₃O₄/FF. (f) XRD patterns of three electrocatalysts.

  DownLoad: Full-Size Img PowerPoint

  To compare the effectiveness in nitrate removal and NH₃ production among the three specially prepared cathodes and a bare FF cathode, the chronoamperometry method was utilized for evaluation. Fig. S4 (Supporting information) reveals that after 6 h of operation at −1.3 V (vs. SCE) with 100 mg/L NO₃⁻-N, the nitrate removal efficiencies were 100%, 56.7%, and 33.6% for Fe@Fe₃O₄/FF, Fe/FF, and Fe₃O₄/CC, respectively. This data indicates the superior efficacy of Fe@Fe₃O₄/FF in NO₃RR, outperforming the other tested cathodes. Of particular note, all NH₃ yield rates were normalized against the electrochemically active surface area (ECSA), which represents the maximum number of available active sites during electrolysis. The ECSA was measured using the well-established double-layer capacitance method, as detailed in Figs. S5 and S6 (Supporting information). As illustrated in Figs. 2a and b, the Fe@Fe₃O₄/FF cathode exhibited significantly higher NH₃ yield performance across all tested potentials compared to its counterparts. It is noticeable that the NH₃ yield rate for Fe@Fe₃O₄/FF increased progressively from 0.056 mmol h⁻¹ cm⁻² to 0.28 mmol h⁻¹ cm⁻² as the potential shifted more negatively from −1.1 V to −1.6 V (vs. SCE), achieving a high FE of 95.7% even at the high overpotential of −1.6 V (vs. SCE). This highlights the exceptional capability of Fe@Fe₃O₄/FF in facilitating NH₃ production via NO₃RR. In contrast, the peak NH₃ yield rates and FE at −1.6 V (vs. SCE) for the other cathodes were significantly lower, recording 0.10 mmol h⁻¹ cm⁻²/53.0%, 0.14 mmol h⁻¹ cm⁻²/82.7%, and 0.076 mmol h⁻¹ cm⁻²/44.1% for the FF, Fe/FF, and Fe₃O₄/CC, respectively, which were all markedly inferior to the performance of Fe@Fe₃O₄/FF. In addition, the NH₃ yield response to varying nitrate concentrations exhibited distinct patterns for Fe/FF, Fe₃O₄/CC, and Fe@Fe₃O₄/FF, as depicted in Fig. 2c, where each catalyst showed a volcano-shaped yield curve, albeit with different nitrate concentration turning points. The NH₃ yield rates for Fe₃O₄/CC and Fe@Fe₃O₄/FF followed a similar trend, increasing with nitrate concentration from 25 mmol/L to 300 mmol/L before showing a decline, with peak NH₃ yields of 0.11 mmol h⁻¹ cm⁻² and 0.23 mmol h⁻¹ cm⁻², respectively. Fe/FF exhibited a decrease in the NH₃ yield rate at a lower nitrate concentration of 100 mmol/L, indicating its higher sensitivity to nitrate concentration. The suppression of NH₃ yield at high nitrate concentrations occurred likely because the nitrite produced could not be continuously reduced to ammonia. This was due to the high concentration of nitrate in the bulk solution, which competed with the intermediate nitrite for electrons on the cathode.

  Figure 2

  

  Figure 2.  Effects of applied potential on (a) NH₃ yield rate and (b) FE of NH₃, and (c) the impact of nitrate concentration on NH₃ yield rate across all the investigated cathodes. Nitrate concentration-dependent FE of various products during NO₃RR for (d) Fe@Fe₃O₄/FF, (e) Fe₃O₄/CC, and (f) Fe/FF. Unless otherwise specified, reaction conditions are as follows: applied potential of −1.3 V (vs. SCE), 100 mmol/L Na₂SO₄, 100 mmol/L NaNO₃, and 1 h of electrolysis.

  DownLoad: Full-Size Img PowerPoint

  As illustrated in Figs. 2d-f, a remarkable increase in the accumulation of the undesired nitrite byproduct was observed for Fe/FF, with the FE(NO₂⁻) values increasing across nitrate concentrations, reaching up to 69.3% at 500 mmol/L nitrate and −1.3 V (vs. SCE). Meanwhile, the maximum FE(NO₂⁻) for Fe@Fe₃O₄/FF and Fe₃O₄/CC was significantly lower, indicating a more controlled nitrite formation during NO₃RR, especially at elevated nitrate concentrations. The values of FE(H₂) resulting from the competitive HER were also examined. It is revealed that this competitive reaction predominantly occurred at lower nitrate concentrations, with the Fe₃O₄/CC cathode more inclined towards favoring HER. Specifically, at 25 mmol/L nitrate, the FE(H₂) values were ranked as follows: Fe₃O₄/CC (44.42%) > Fe@Fe₃O₄/FF (13%) > Fe/FF (2%). Overall, the FE(NH₃) of Fe@Fe₃O₄/FF outperformed Fe₃O₄/CC and Fe/FF under all tested conditions, highlighting its superior ability for the selective production of NH₃.

  Comprehensive evidence was presented to elucidate the mechanism underlying the exceptional performance of Fe@Fe₃O₄/FF in NO₃RR, particularly its remarkable NH₃ yield and substantially decreased nitrite production. The analysis involved three cathodes, Fe@Fe₃O₄/FF, Fe/FF, and Fe₃O₄/CC, and a control electrode, Fe₃O₄/nickel foam (NF), which was prepared in a manner analogous to Fe₃O₄/CC but utilized NF as the substrate. Figs. 3a-c and Fig. S7 (Supporting information) show the effects of tertiary butanol (TBA), a recognized scavenger of ^•H, on the linear sweep voltammetry (LSV) responses of various electrodes. The differences in current density reduction observed in the LSV curves upon the addition of TBA implied the varied influence of ^•H in NO₃RR among various electrodes. Specifically, Fe@Fe₃O₄/FF exhibited a reduction in current density from 12.32 mA/cm² to 10.93 mA/cm² at −1.3 V (vs. SCE) in the presence of TBA, indicating a current density inhibition of 10.1%. Meanwhile, the suppression ratios of current density after TBA addition for the other electrodes were significantly higher, at 45.2% for Fe/FF, 77.4% for Fe₃O₄/CC, and 84.7% for Fe₃O₄/NF, respectively. This differential response suggests a predominant role of ^•H in the NO₃RR for these electrodes, while DET may primarily facilitate NO₃RR on Fe@Fe₃O₄/FF.

  Figure 3

  

  Figure 3.  Current density variations in LSV for (a) Fe@Fe₃O₄/FF, (b) Fe/FF, and (c) Fe₃O₄/CC electrodes following the addition of 300 mmol/L TBA and 100 mg/L NaNO₃. (d) Impact of TBA concentration on the nitrate removal efficiency across four electrodes, and (e) corresponding calculations of the contributions from different mechanisms involved in NO₃RR. (f) Correlation between the NH₃ yield rate and the proportion of DET in NO₃RR, as determined from (e). (g) Measurement of the OH⁻ adsorptive peak in 1 mol/L NaOH solution. (h) EIS spectra of the three electrodes at −1.1 V (vs. SCE) in the presence of 100 mg/L NaNO₃. (i) Influence of different alkali metal cations in the electrolyte on the NH₃ yield rate for the three electrodes, evaluated in a solution of 100 mmol/L M₂SO₄ and 100 mmol/L MNO₃ at −1.3 V (vs. SCE).

  DownLoad: Full-Size Img PowerPoint

  Further investigation into the reaction mechanisms focused on evaluating nitrate removal efficiency in the presence of varying concentrations of TBA. Fig. 3d shows that Fe@Fe₃O₄/FF had minimal sensitivity to changes in TBA concentration, with removal efficiency slightly diminishing from 99.5% to 80.2% as the molar ratio of TBA to nitrate (C_(TBA)/C_(N)) was elevated from 0 to 50. This modest decline suggests a relatively low ^•H contribution of 19.4% to the NO₃RR process for Fe@Fe₃O₄/FF, as illustrated in Fig. 3e. In contrast, the other electrodes experienced a more marked decrease in nitrate removal efficiency with the same increase in TBA concentration, with efficiencies dropping to 59.9%, 38.0%, and 19.5% for Fe/FF, Fe₃O₄/CC, and Fe₃O₄/NF, respectively. This pattern, coupled with the distinct reduction in current density observed in LSV curves after TBA was added, infers ^•H contributions of 38.5%, 60.5%, and 88.6% to NO₃RR for Fe/FF, Fe₃O₄/CC, and Fe₃O₄/NF, respectively. In addition, after performing reactions with a C_(TBA)/C_(N) ratio of 50 for all electrodes, the nitrate removal efficiency was reevaluated without the addition of TBA. Interestingly, the reassessment resulted in efficiency values that were almost identical to those recorded in the initial tests under the same conditions for all electrodes. This consistency indicates that the inhibitory effect of TBA stems from its scavenging activity rather than a poisoning effect on the electrodes [25,26]. This conclusion highlights the distinctive pathways through which each electrode facilitates NO₃RR, with Fe@Fe₃O₄/FF exhibiting a particularly efficient and less ^•H-dependent reaction mechanism.

  The ESR technology was utilized to semi-quantitatively measure the amount of electrochemically generated ^•H across the four electrodes under identical conditions. As illustrated in Fig. S8 (Supporting information), the characteristic peaks for 5,5-dimethyl-1-pyrroline N-oxide (DMPO)-^•H, a marker for ^•H, were detected for all four electrodes, albeit with varying signal intensities. Fe₃O₄/NF exhibited the highest peak intensity of DMPO-^•H, followed by Fe₃O₄/CC, Fe/FF, and Fe@Fe₃O₄/FF, respectively. This order aligns with the previously determined ^•H contributions to the NO₃RR, again delineating the different predominant reaction mechanisms for each electrode. Combining these findings with earlier experimental results, it is conclusively determined that Fe@Fe₃O₄/FF primarily facilitates NO₃RR via the DET mechanism. In contrast, HAT predominantly drives the reaction for Fe₃O₄/CC and Fe₃O₄/NF, while both DET and HAT mechanisms are active in Fe/FF. Intriguingly, when correlating the NH₃ yield rate via NO₃RR at −1.3 V (vs. SCE) with the contribution of the DET mechanism, a nearly linear relationship appeared (Fig. 3f). This suggests that the superior NH₃ yield rate and substantially reduced nitrite production observed for Fe@Fe₃O₄/FF may be attributed to the dominance of the DET mechanism and minimal involvement of HAT. Therefore, a significant contribution from the DET mechanism is fundamentally crucial for the exceptional performance of Fe@Fe₃O₄/FF in achieving high NH₃ generation.

  Based on the established relationship between reaction mechanisms and NO₃RR performance, the next step was to elucidate how the inherent properties of electrocatalysts influence these mechanisms. It is hypothesized that the strong adsorption of NO₃⁻ and other nitrogen-containing intermediates onto the surface of electrocatalysts is crucial for the NO₃RR to proceed via the DET mechanism [20,27,28], thereby leading to enhanced NO₃RR performance. To test this hypothesis, the adsorption of OH⁻ serving as a proxy for NO₃⁻ and other nitrogen-containing intermediates generated during NO₃RR was measured using the cyclic voltammetry (CV) technology. The results in Fig. 3g show that Fe@Fe₃O₄/FF exhibited the strongest OH⁻ adsorption, as indicated by the most negative potential of the OH⁻ adsorptive peak at −0.813 V (vs. SCE). This suggests a potentially strong bond between nitrogen-containing species and Fe@Fe₃O₄/FF, facilitating direct electron transfer. In comparison, the OH⁻ adsorptive peaks for Fe/FF and Fe₃O₄/CC were observed at −0.774 and −0.755 V (vs. SCE), respectively, indicating a better adsorption affinity for nitrogen-containing species on Fe/FF than on Fe₃O₄/CC, but still inferior to Fe@Fe₃O₄/FF. No peak corresponding to OH⁻ adsorption was detected on Fe₃O₄/NF, indicating inferior nitrate adsorption for Fe₃O₄/NF. The measured order of OH⁻ adsorptive peak potentials correlated well with the contribution of the DET mechanism in NO₃RR for the electrodes, suggesting that enhanced adsorption of nitrogen-containing species is favorable for NO₃RR via the DET mechanism. Moreover, the interfacial charge transfer resistance for Fe@Fe₃O₄/FF, Fe/FF, Fe₃O₄/CC and Fe₃O₄/NF was determined using electrochemical impedance spectroscopy (EIS) (Fig. 3h). As anticipated, Fe@Fe₃O₄/FF displayed the lowest resistance to charge transfer, as evidenced by the smallest diameter of the corresponding semi-circle in the EIS spectra. This result supports the most favorable interfacial electron transfer for Fe@Fe₃O₄/FF, followed by Fe/FF, Fe₃O₄/CC and Fe₃O₄/NF, and aligns with the previously mentioned contribution of the DET mechanism in NO₃RR.

  The role of protonation kinetics, alongside enhanced adsorption of nitrogen-containing species, is also crucial for achieving high NH₃ yield, especially given the nature of electrochemical hydrogenation in aqueous media. This study examined the protonation kinetics of three electrocatalysts by altering the alkali metal cation of the electrolyte from Li⁺ to Na⁺, K⁺, and Cs⁺, a method thoroughly documented in existing literature [29-31]. Fig. 3i shows that Fe@Fe₃O₄/FF, Fe₃O₄/CC, and Fe₃O₄/NF all exhibited progressively improved NH₃ yield rates with the transition from Li⁺ to Cs⁺, increasing from 0.074, 0.03, and 0.0058 mmol h⁻¹ cm⁻² to 0.15, 0.08, and 0.02 mmol h⁻¹ cm⁻², respectively. This improvement indicates the favorable protonation kinetics of Fe@Fe₃O₄/FF, Fe₃O₄/CC, and Fe₃O₄/NF, consistent with trends reported previously [30,32]. In contrast, Fe/FF demonstrated a minimal improvement in NH₃ yield rate, from mmol h⁻¹ cm⁻² 0.03 to 0.04 mmol h⁻¹ cm⁻² under the same conditions, indicating weaker protonation kinetics and manifesting the critical role of Fe₃O₄ in enhancing NO₃RR protonation. Interestingly, the superior protonation kinetics of Fe@Fe₃O₄/FF and Fe₃O₄/CC were closely associated with a significantly reduced accumulation of nitrite, as observed in Figs. 2d-f. This suggests that the presence of Fe₃O₄ in these electrocatalysts facilitates the continuous protonation of intermediate nitrite into NH₃. This effect is likely due to the inherent ability of common metal oxides, such as Fe₃O₄, to adsorb and dissociate water effectively [33,34].

  DFT calculations were employed to deepen the understanding of the relationship between catalyst structure and NO₃RR performance, particularly focusing on the advantages of the DET mechanism over the HAT mechanism. Three representative models (Fig. 4a), Fe(211) for Fe/FF, Fe₃O₄(220) for Fe₃O₄/CC, and Fe@Fe₃O₄(220) for Fe@Fe₃O₄/FF, were developed based on the experimental characterizations presented in Fig. 1, to closely reflect the structures of the actual catalysts. The adsorption free energy of NO₃⁻ on these models was calculated to identify which structure favored NO₃⁻ binding, a significant aspect in determining the reaction mechanism [35]. The results, shown in Fig. 4b, indicate that Fe@Fe₃O₄(220) exhibited the highest ΔG for NO₃⁻ adsorption at −0.772 eV, followed by −0.526 eV for Fe(221) and −0.415 eV for Fe₃O₄(220). This suggests that the Fe component in Fe@Fe₃O₄/FF plays a pivotal role in enhancing NO₃⁻ adsorption, thereby contributing to its superior NO₃RR performance. The slightly stronger NO₃⁻ adsorption on Fe@Fe₃O₄(220) compared to Fe(221) also highlighted the beneficial effect of the coexisting Fe₃O₄ in Fe@Fe₃O₄/FF, aligning with the experimentally observed trends in NO₃⁻ adsorption. In addition, the kinetic energy barrier for the initial step of NO₃RR (NO₃⁻ → NO₂⁻), often considered the rate-limiting step [36], was evaluated for both HAT and DET mechanisms. The HAT mechanism involves the direct hydrogenation of adsorbed NO₃⁻ (*NO₃) by adjacent ^•H, while DET facilitates this transformation through sequential proton and electron transfer to adsorbed NO₃⁻ without the involvement of neutral ^•H. The energy profiles for NO₃⁻ transformation into NO₂⁻ + OH⁻ via both mechanisms were analyzed for all three models (Figs. 4c and d). Notably, Fe₃O₄(220) exhibited the lowest energy barrier of 1.01 eV for the HAT mechanism, followed by 1.28 eV for Fe(221) and 1.70 eV for Fe@Fe₃O₄(220), consistent with the experimentally measured contributions of the HAT mechanism. Conversely, Fe@Fe₃O₄(220) presented the lowest energy barrier of 0.42 eV for the DET mechanism, followed by 0.63 eV for Fe(221) and 0.99 eV for Fe₃O₄(220), theoretically supporting the preference for NO₃RR via DET on Fe@Fe₃O₄/FF and suggesting a significant contribution of the HAT mechanism for NO₃RR over Fe₃O₄/CC. These theoretical outcomes confirm the DET mechanism's favorable position relative to the HAT mechanism, especially since Fe@Fe₃O₄/FF, dominated by DET, exhibited the best nitrate conversion capability among the three investigated electrodes. This comprehensive analysis not only corroborates the experimental observations but also provides a mechanistic understanding of why Fe@Fe₃O₄/FF outperforms its counterparts in NO₃RR, highlighting the critical role of catalyst structure in dictating the reaction pathway and efficiency.

  Figure 4

  

  Figure 4.  (a) Representations of Fe/FF, Fe₃O₄/CC, and Fe@Fe₃O₄/FF models used for theoretical calculations. (b) Comparative analysis of NO₃⁻ adsorption free energy across the three models. Evaluation of the kinetic energy barrier for (c) HAT and (d) DET mechanisms on each of the models.

  DownLoad: Full-Size Img PowerPoint

  The collaborative experimental and theoretical analyses provided profound insights into the mechanics and effects of different components within electrocatalysts on NO₃RR. Fig. 5 encapsulates these insights, highlighting how the interplay between metallic Fe and Fe₃O₄ within Fe@Fe₃O₄/FF enhances the NO₃RR process. The role of metallic Fe is pivotal due to its strong adsorption capabilities for nitrogen-containing species, facilitated by intensive D-π interactions between the electron-rich metallic Fe surface and N atoms [37]. This interaction fosters a preference for the DET mechanism in NO₃RR, attributed to reduced interfacial resistance for charge transfer. Concurrently, the inclusion of Fe₃O₄ in Fe@Fe₃O₄/FF is advantageous for promoting favorable protonation kinetics, thanks to its high affinity for water adsorption and dissociation. This is attributed to the strong interaction between the empty d orbitals of the metal oxide and the lone-pair electrons of the oxygen atom in water molecules [33,38]. This characteristic significantly mitigates nitrite accumulation and boosts NH₃ production. In contrast, while Fe/FF also demonstrates robust nitrogen-containing species binding and low interfacial charge transfer resistance, its sole reliance on metallic Fe without the presence of Fe₃O₄ limits its protonation kinetics. This limitation could lead to considerable nitrite accumulation, especially at higher nitrate concentrations. Fe₃O₄/CC, characterized by effective protonation kinetics, however, exhibits poor NH₃ yield and slightly reduced nitrite accumulation, primarily influenced by the HAT mechanism. Therefore, the synergy between Fe and Fe₃O₄ in Fe@Fe₃O₄/FF is crucial for achieving superior NH₃ yield through the DET mechanism. This insight is useful for the design and development of highly effective electrocatalysts for NO₃RR, emphasizing the need to balance components that enhance nitrogen-containing species adsorption and promote efficient protonation kinetics to minimize undesirable byproducts and maximize desired outcomes.

  Figure 5

  

  Figure 5.  Schematic representation of reaction mechanisms on (a) Fe@Fe₃O₄/FF, (b) Fe/FF, and (c) Fe₃O₄/CC.

  DownLoad: Full-Size Img PowerPoint

  The performance of Fe@Fe₃O₄/FF for complete denitrification under various conditions, including the presence of NaCl, was thoroughly assessed. The addition of NaCl aimed to initiate electrochemically mediated breakpoint chlorination, oxidizing the NH₄⁺ generated from NO₃RR into harmless N₂ [39-42]. Fig. S9a (Supporting information) shows that 500 mg/L NO₃⁻-N can be fully removed at 30 mA/cm² after 3 h of electrolysis, achieving a pseudo-first-order rate constant of 1.39 h⁻¹. This rate constant was observed to increase incrementally to 1.43, 1.49, and 1.52 h⁻¹ as the current density was raised to 35, 40, and 45 mA/cm², respectively. Negligible amounts of NH₄⁺-N and NO₂⁻-N were detected after the electrolysis process, indicating the capability of Fe@Fe₃O₄/FF for complete denitrification with the assistance of NaCl.

  The impacts of initial NO₃⁻-N concentration and NaCl concentration on the denitrification efficiency of Fe@Fe₃O₄/FF were explored, as presented in Figs. S9b and c (Supporting information). An observed trend showed a gradual decline in the kinetics as the initial NO₃⁻-N concentration increased from 500 mg/L to 800 mg/L, attributed to the limitation in active sites [43]. Nevertheless, nearly 100% denitrification efficiency was still achieved after 3 h of electrolysis, even at the elevated NO₃⁻-N concentration of 800 mg/L. Complete removal of 500 mg/L NO₃⁻-N was observed, irrespective of the NaCl amount. The denitrification kinetics exhibited a slight improvement with increasing concentrations of NaCl. This enhancement may be attributed to the refreshment of active sites occupied by the NH₃ product and the alleviation of pH increase in the solution through electrochemically mediated breakpoint chlorination. This suggests a strong resistance to oxidative corrosion by active chlorine species produced from Cl⁻ oxidation at the anode [41,42]. The Fe@Fe₃O₄/FF cathode maintained exceptional denitrification efficiency across a pH range from 7 to 12 (Fig. S9d in Supporting information), suggesting the promising potential for practical applications. The acidic conditions were not examined in the present study due to inevitable dissolution of Fe₃O₄ under these conditions. Material stability, a critical performance metric, was evaluated over 20 cycles. As shown in Fig. S10 (Supporting information), the Fe@Fe₃O₄/FF cathode showed consistent denitrification efficiency of approximately 100% removal across all cycles, indicating its exceptional robustness. The post-XRD pattern of Fe@Fe₃O₄/FF in Fig. S11 (Supporting information) displayed a profile similar to that of the pristine Fe@Fe₃O₄/FF. These findings highlight the effectiveness and durability of Fe@Fe₃O₄/FF in complex water treatment scenarios, offering a promising approach to achieving complete denitrification in the presence of challenging constituents like NaCl.

  Given the advantages of Fe@Fe₃O₄/FF, such as cost-effectiveness and ease of scale-up, the practical feasibility of using Fe@Fe₃O₄/FF as a cathode material for treating real wastewater with high nitrate concentrations was investigated. To facilitate complete denitrification through electrochemically mediated breakpoint chlorination, an additional 100 mmol/L NaCl was supplemented into the wastewater. An experimental setup featuring an internal circulation mode was constructed, as depicted in Fig. S12a (Supporting information). This setup was operated in an internal-circulation mode, including a peristaltic pump for flow rate control, an electrolyzer equipped with a larger-sized Fe@Fe₃O₄/FF cathode (4 cm × 3 cm) for denitrification, and a container for storing and mixing 500 mL of the original wastewater with the effluent from the electrolyzer. Operating at a current density of 30 mA/cm² and a flow rate of 1 mL/min, the NO₃⁻-N concentration in the wastewater was steadily reduced from 3815 mg/L to zero over 120 h of electrolysis, as shown in Fig. S12b (Supporting information). The byproducts of NH₄⁺-N and NO₂⁻-N were scarcely accumulated by the end of the process, demonstrating the effectiveness of active chlorine species generated from Cl⁻ oxidation at the MMO anode in ensuring efficient denitrification. In stark contrast, when using Fe/FF as the cathode under identical operating conditions, the denitrification performance was significantly inferior, achieving only a 29.3% removal efficiency.

  In conclusion, through detailed investigations of NO₃RR on Fe/FF, Fe₃O₄/CC, and Fe@Fe₃O₄/FF cathodes, we have clearly established that the catalyst structure directly influences the reaction mechanism and, consequently, the NO₃RR performance. The Fe@Fe₃O₄/FF cathode demonstrated a strong affinity for NO₃⁻ and a moderate capability for water dissociation, being attributed to its metallic Fe and Fe₃O₄ components, respectively. These characteristics are key to the preferred occurrence of the DET mechanism during NO₃RR, leading to the outstanding denitrification performance of Fe@Fe₃O₄/FF. Conversely, the absence of metallic oxide Fe₃O₄ in Fe/FF, which is crucial for the protonation process, resulted in poor NO₃RR performance and significant accumulation of the intermediate product nitrite. Fe₃O₄/CC, containing only the metallic oxide Fe₃O₄ component, exhibited inferior denitrification performance. The lack of metallic Fe, which is essential for strong bonding with NO₃⁻, leads to a preference for the HAT mechanism during NO₃RR. This preference accelerates the competitive HER and limits denitrification performance, despite a noticeable reduction in the accumulation of NO₂⁻ due to the enhanced protonation capability of Fe₃O₄. These experimental insights are crucial for a fundamental understanding of the DET and HAT mechanisms involved in NO₃RR. They also provide a solid foundation for the rational design and development of advanced catalysts for effective nitrate reduction, highlighting the importance of carefully considering catalyst composition and structure to optimize performance.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Yuwei Liu: Writing – original draft, Investigation, Formal analysis, Data curation. Yihui Zhu: Investigation, Formal analysis, Data curation. Weijian Duan: Writing – original draft, Methodology, Data curation, Conceptualization. Yizhuo Yang: Resources, Data curation. Haorui Tuo: Investigation, Data curation. Chunhua Feng: Writing – review & editing, Supervision, Methodology, Formal analysis, Conceptualization.

  Acknowledgments

  We gratefully acknowledge financial support from the National Natural Science Foundation of China (Nos. U21A2034 and 21876052), the Guangdong Special Support Plan for Innovation Teams (No. 2019BT02L218), the Guangdong Special Support Plan for Young Top-notch Talents (No. 2019TQ05L179), and the Natural Science Foundation of Guangdong Province, China (No. 2021B1515120077).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110347.

  
  
• 1. [1]

     M. Duca, M.T.M. Koper, Energy Environ. Sci. 5 (2012) 9726–9742. doi: 10.1039/c2ee23062c

  2. [2]

     H. Xu, Y. Ma, J. Chen, W.X. Zhang, J. Yang, Chem. Soc. Rev. 51 (2022) 2710–2758. doi: 10.1039/d1cs00857a

  3. [3]

     Y. Xiong, Y. Wang, J. Zhou, et al., Adv. Mater. 36 (2024) 2304021. doi: 10.1002/adma.202304021

  4. [4]

     Y. Wang, C. Wang, M. Li, Y. Yu, B. Zhang, Chem. Soc. Rev. 50 (2021) 6720–6733. doi: 10.1039/d1cs00116g

  5. [5]

     W. Zheng, L. Zhu, Z. Yan, et al., Environ. Sci. Technol. 55 (2021) 13231–13243.

  6. [6]

     T. Zhu, Q. Chen, P. Liao, et al., Small 16 (2020) 2004526. doi: 10.1002/smll.202004526

  7. [7]

     Y. Ren, S. You, Y. Wang, J. Yang, Y. Liu, Environ. Sci. Technol. 58 (2024) 2144–2152. doi: 10.1021/acs.est.3c09759

  8. [8]

     J. Guo, P. Chen, Chem 3 (2017) 709–712. doi: 10.1016/j.chempr.2017.10.004

  9. [9]

     J. Li, G. Zhan, J. Yang, et al., J. Am. Chem. Soc. 142 (2020) 7036–7046. doi: 10.1021/jacs.0c00418

  10. [10]

      M. Fu, Y. Mao, H. Wang, et al., Chin. Chem. Lett. 35 (2024) 108341. doi: 10.1016/j.cclet.2023.108341

  11. [11]

      Y. Xu, K. Shi, T. Ren, et al., Small 18 (2022) 2203335. doi: 10.1002/smll.202203335

  12. [12]

      Y. Wang, A. Xu, Z. Wang, et al., J. Am. Chem. Soc. 142 (2020) 5702–5708. doi: 10.1021/jacs.9b13347

  13. [13]

      H. Yin, X. Mao, S. Bell, D. Golberg, A. Du, Chem. Mater. 35 (2023) 2884–2891. doi: 10.1021/acs.chemmater.2c03788

  14. [14]

      X. Fan, L. Xie, J. Liang, et al., Nano Res. 15 (2022) 3050–3055. doi: 10.1007/s12274-021-3951-5

  15. [15]

      L. Fang, S. Wang, S. Lu, et al., Chin. Chem. Lett. 35 (2024) 108864. doi: 10.1016/j.cclet.2023.108864

  16. [16]

      Z.Y. Wu, M. Karamad, X. Yong, et al., Nat. Commun. 12 (2021) 2870. doi: 10.1038/s41467-021-23115-x

  17. [17]

      X.F. Cheng, J.H. He, H.Q. Ji, et al., Adv. Mater. 34 (2022) 2205767. doi: 10.1002/adma.202205767

  18. [18]

      Y. Xue, Q. Yu, Q. Ma, et al., Environ. Sci. Technol. 56 (2022) 14797–14807. doi: 10.1021/acs.est.2c04456

  19. [19]

      J. Gao, B. Jiang, C. Ni, et al., Appl. Catal. B 254 (2019) 391–402. doi: 10.1016/j.apcatb.2019.05.016

  20. [20]

      Y. Li, J. Ma, Z. Wu, Z. Wang, Environ. Sci. Technol. 56 (2022) 8673–8681. doi: 10.1021/acs.est.1c05841

  21. [21]

      X. Li, P. Shen, X. Li, D. Ma, K. Chu, ACS Nano 17 (2023) 1081–1090. doi: 10.1021/acsnano.2c07911

  22. [22]

      N. Zhang, G. Zhang, P. Shen, et al., Adv. Funct. Mater. 33 (2023) 2211537. doi: 10.1002/adfm.202211537

  23. [23]

      K. Fan, W. Xie, J. Li, et al., Nat. Commun. 13 (2022) 7958. doi: 10.1038/s41467-022-35664-w

  24. [24]

      W. Duan, Y. Chen, H. Ma, et al., Environ. Sci. Technol. 57 (2023) 3893–3904. doi: 10.1021/acs.est.2c09147

  25. [25]

      R. Mao, N. Li, H. Lan, et al., Environ. Sci. Technol. 50 (2016) 3829–3837. doi: 10.1021/acs.est.5b05006

  26. [26]

      H. Lan, R. Mao, Y. Tong, et al., Environ. Sci. Technol. 50 (2016) 11872–11878. doi: 10.1021/acs.est.6b02822

  27. [27]

      W. Fu, Y. Du, J. Jing, C. Fu, M. Zhou, Appl. Catal. B 324 (2023) 122201. doi: 10.1016/j.apcatb.2022.122201

  28. [28]

      X. Zhao, Y. Jiang, M. Wang, et al., Adv. Energy Mater. 13 (2023) 2301409. doi: 10.1002/aenm.202301409

  29. [29]

      S. Xue, B. Garlyyev, S. Watzele, et al., ChemElectroChem 5 (2018) 2326–2329. doi: 10.1002/celc.201800690

  30. [30]

      W. Ma, S. Xie, T. Liu, et al., Nat. Catal. 3 (2020) 478–487. doi: 10.1038/s41929-020-0450-0

  31. [31]

      X. Lu, W. Tu, Y. Zhou, Z. Zou, Adv. Energy Mater. 13 (2023) 2300628. doi: 10.1002/aenm.202300628

  32. [32]

      W. Ma, S. Xie, X.G. Zhang, et al., Nat. Commun. 10 (2019) 892. doi: 10.1038/s41467-019-08805-x

  33. [33]

      H. Sun, Z. Yan, C. Tian, et al., Nat. Commun. 13 (2022) 3857. doi: 10.1038/s41467-022-31561-4

  34. [34]

      K.L. Zhou, Z. Wang, C.B. Han, et al., Nat. Commun. 12 (2021) 3783. doi: 10.1038/s41467-021-24079-8

  35. [35]

      Y. Hua, N. Song, Z. Wu, et al., Adv. Funct. Mater. 34 (2024) 2314461. doi: 10.1002/adfm.202314461

  36. [36]

      H. Jiang, G.F. Chen, O. Savateev, et al., Wang, Angew. Chem. Int. Ed. 62 (2023) 18717.

  37. [37]

      Y. Dong, J.B. Chen, J. Ying, et al., Chem. Mater. 34 (2022) 8271–8279. doi: 10.1021/acs.chemmater.2c01738

  38. [38]

      W. Shen, Y. Zheng, Y. Hu, et al., J. Am. Chem. Soc. 146 (2024) 5324–5332. doi: 10.1021/jacs.3c11861

  39. [39]

      Z. Yan, W.J. Kuang, Y. Lei, et al., Environ. Sci. Technol. 57 (2023) 20915–20928. doi: 10.1021/acs.est.3c06326

  40. [40]

      C. Zhang, D. He, J. Ma, T.D. Waite, Water Res. 145 (2018) 220–230. doi: 10.1016/j.watres.2018.08.025

  41. [41]

      F. Li, L. Sun, Y. Liu, et al., J. Hazard. Mater. 400 (2020) 123246. doi: 10.1016/j.jhazmat.2020.123246

  42. [42]

      F. Li, X. Peng, Y. Liu, et al., Chemosphere 229 (2019) 383–391. doi: 10.1016/j.chemosphere.2019.04.180

  43. [43]

      Z.Y. Wu, P. Zhu, D.A. Cullen, et al., Nat. Synth. 1 (2022) 658–667. doi: 10.1038/s44160-022-00129-x

• 

  Figure 1  SEM images of (a) Fe@Fe₃O₄/FF, (b) Fe₃O₄/CC, and (c) Fe/FF. (d, e) HRTEM images of Fe@Fe₃O₄/FF. (f) XRD patterns of three electrocatalysts.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Effects of applied potential on (a) NH₃ yield rate and (b) FE of NH₃, and (c) the impact of nitrate concentration on NH₃ yield rate across all the investigated cathodes. Nitrate concentration-dependent FE of various products during NO₃RR for (d) Fe@Fe₃O₄/FF, (e) Fe₃O₄/CC, and (f) Fe/FF. Unless otherwise specified, reaction conditions are as follows: applied potential of −1.3 V (vs. SCE), 100 mmol/L Na₂SO₄, 100 mmol/L NaNO₃, and 1 h of electrolysis.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Current density variations in LSV for (a) Fe@Fe₃O₄/FF, (b) Fe/FF, and (c) Fe₃O₄/CC electrodes following the addition of 300 mmol/L TBA and 100 mg/L NaNO₃. (d) Impact of TBA concentration on the nitrate removal efficiency across four electrodes, and (e) corresponding calculations of the contributions from different mechanisms involved in NO₃RR. (f) Correlation between the NH₃ yield rate and the proportion of DET in NO₃RR, as determined from (e). (g) Measurement of the OH⁻ adsorptive peak in 1 mol/L NaOH solution. (h) EIS spectra of the three electrodes at −1.1 V (vs. SCE) in the presence of 100 mg/L NaNO₃. (i) Influence of different alkali metal cations in the electrolyte on the NH₃ yield rate for the three electrodes, evaluated in a solution of 100 mmol/L M₂SO₄ and 100 mmol/L MNO₃ at −1.3 V (vs. SCE).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Representations of Fe/FF, Fe₃O₄/CC, and Fe@Fe₃O₄/FF models used for theoretical calculations. (b) Comparative analysis of NO₃⁻ adsorption free energy across the three models. Evaluation of the kinetic energy barrier for (c) HAT and (d) DET mechanisms on each of the models.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Schematic representation of reaction mechanisms on (a) Fe@Fe₃O₄/FF, (b) Fe/FF, and (c) Fe₃O₄/CC.

  下载: 全尺寸图片 幻灯片
•