The electric potential is the main driver of electrochemical nitrate reduction and significantly affects the rate and products of nitrate reduction [37]. In kinetic perspective, electrochemical reduction of nitrate involves reactant diffusion, adsorption, charge transfer, product desorption and separation [21]. Polatides and Kyriacou [38] studied the kinetics of nitrate reduction reaction under Sn₈₅Cu₁₅ cathode and found that the increase in potential favored the formation of N₂ and NH₃. The rate of ammonia production increased more strongly when the potential was below –1.8 V (vs. Ag/AgCl). It is thought to require hydrogenation of the adsorbed NO for ammonia formation at high potentials, thus facilitating the rate of ammonia production with the increase in surface coverage of adsorbed hydrogen. Similarly, the CuO@PANI/NF cathode achieved the highest NO₃⁻–N conversion efficiency (97.16%) at –1.3 V (vs. SCE). At this potential, the NH₃ yield, FE and selectivity all reached their peak (0.213 mmol h⁻¹ cm⁻², 93.88% and 91.38%). Study at copper electrodes reduced the potential from –1.1 V to –1.2 V and then to –1.4 V (vs. Hg/HgO). The selectivity of ammonium first increased from 13% to 39% and then increased rapidly to 80% [39]. It was thought to be an inconsistent electron requirement for the different products, while Eqs. 2 and 3 also suggest that more negative potentials favor the production of ammonia. However, when the potential is too low, nitrate reduction is inhibited. When the potential decreased from –1.8 V to –2.0 V (vs. SCE), the nitrate reduction rate decreased from 15% to 8% after 120 min of electrolysis on the graphite cathode [40]. It was considered to be the effect of side reactions (e.g., Hydrogen evolution reaction (HER)). The In-S-G catalyst exhibited a maximum NH₃ yield of 375 mmol g_cat⁻¹ h⁻¹ at –0.7 V (vs. RHE). The NH₃ yield remained around 321 mmol g_cat⁻¹ h⁻¹ when the potential dropped to –1.0 V (vs. RHE). However, the FE decreased from the peak of 75% (–0.5 V vs. RHE) to 36%. It was also thought to be caused by the competing processes of the HER at high potentials. In summary, the potential for studying the electrochemical nitrate production of ammonia usually lies between –0.8 V to –1.8 V (vs. SCE). The occurrence of side reactions can be suppressed to the maximum extent within the range, thus improving the selectivity of ammonia. Therefore, the optimal ammonia generation potential exhibited by different cathodes varies depending on the adsorption location and potential region of the material.

In the field of wastewater treatment, there is a wide variety of nitrate-rich wastewaters and a wide range of nitrate concentrations. The nitrate concentration in municipal wastewater is mostly 10–200 mg/L NO₃⁻–N, which is the prevailing range in the current research on electrochemical nitrate to ammonia [41]. For example, Pd Cu(OH)₂/CF as a cathode could produce up to 98.8% ammonium selectivity at 50 mg/L NO₃⁻ [42]. In contrast, under a higher nitrate concentration (500 mg/L NO₃⁻), the ammonia production rate and efficiency also could reach 436 ± 85 µg h⁻¹ cm⁻² and 97.8%, respectively, using copper-incorporated crystalline 3,4,9,10-perylenetracarboxylic dianhydride as the cathode material [43]. Some cathode materials with commercial application potential (e.g., graphite felt and BDD) also exhibited impressive nitrate removal (70% and > 90%) in this NO₃⁻–N concentration range (65–500 mg/L) [40,44].

Of course, many industries also produce high concentrations of nitrate-rich wastewater, such as cellophane, explosives, fertilizer, pectin and metal finishing industries, where the nitrate concentration even exceeds 1000 mg/L NO₃⁻–N [41]. Similarly, electrochemical nitrate production of ammonia can also be accomplished at this concentration. Co-NAs catalysts achieved ammonia Faradaic efficiency of over 96% with the NO₃⁻–N concentration of 2800 mg/L [6]. 14000 mg/L NO₃⁻–N achieved 93%-98% faradaic efficiencies towards NH₃ production by ENRA with Pt_xRu_y [45].

Nitrate concentrations often directly affect the kinetic sequence of electrochemical reduction and the activation point of the catalyst surface. It is a conjecture based on the kinetic equations of nitrate reduction by Langmuir (Eq. 18), Katsounaros and Kyriacou (Eq. 19) [46].

where K (= k₁/k_-1) is the ionic association constant, indicating that nitrate tends to form ion pairs. The calculation of Eq. 19 shows that 1 + K[NO₃⁻] ≈ K[NO₃⁻] in a high nitrate concentration environment and the reduction reaction follows zero-order kinetics. However, 1 + K[NO₃⁻] ≈ 1 when the nitrate concentration is low enough, the first-order kinetics becomes the rule to be followed. To prove the conclusion, 62,000 mg/L (1 mol/L) and 6200 mg/L (0.1 mol/L) NO₃⁻ were used as conditions, respectively, with the former following zero-order kinetics within the first 75 min of the reaction, while the latter exhibited first-order kinetics [46]. The phenomenon may be attributed to the influence of intermediates and other ions, resulting in competition for the catalytic position on the electrode surface, where nitrate no longer dominates and thus the reduction rate decreases [47]. Therefore, nitrate is far more competitive than other ions when the concentration is greater than 62,000 mg/L (1 mol/L) NO₃⁻, the reaction rate is almost unaffected. For example, the Strained Ru nanoclusters material was used to reduce 62,000 mg/L NO₃⁻ wastewater with ammonia selectivity close to 100% [48]. However, this type of wastewater is not particularly common, only nitric acid cleaning wastewater from nuclear weapons production plants and research laboratories for radioactive metal products, which can contain more than 50,000 mg/L NO₃⁻–N [41]. Thus, high concentration has more potential when treating nitrate-rich wastewater for electrochemical ammonia production.

Significantly, the concentration and type of electrolyte affect the reduction of nitrate [49]. A portion of industrial and municipal wastewater has lower conductivity, even as low as 1 mS/cm [50]. Low conductivity tends to cause high charge transfer impedance of the system, which brings unnecessary economic losses. Therefore, the addition of salt and buffering agents (Na₂SO₄, NaCl, phosphate buffer saline) or application to high conductivity wastewater (landfill leachate) can effectively alleviate this defects [51].

In fact, the inorganic ions in wastewater are so complicated, such as SO₄²⁻, Cl⁻, Ca²⁺, Mg²⁺, Na⁺ and K⁺. For example, cations can also act as "cation catalysts". The cation forms an instantaneous neutral ion pair by changing the bilayer structure of the cathode. And the ion pair can act as an attraction center for the reducing anion. The repulsion between anion and cathode is then weakened, thus aiding the reduction of nitrate at the cathode [52]. It was found that cations increase the rate of nitrate reduction along the trend of Li⁺ < Na⁺ < K⁺ < Cs⁺ [53]. Even higher rates than alkali metals can be obtained in the presence of multivalent cations such as NH₄⁺, Ca²⁺ and La³⁺ [54]. On the other hand, the cation also has a certain influence on the product. Manzo-Robledo, Lévy-Clément and Alonso-Vante [55] found that the presence of K⁺ makes N₂ the dominant product, while N₂O dominates in the presence of Na⁺. HER was more difficult to occur in K⁺-containing electrolyte solutions compared to Na⁺, so the production of N₂ was less inhibited. The key factor is the adsorption of nitrogen in this study. It seems that the production of HER and nitrogen species is regulated by co-adsorption of alkali cations in the presence of alkali metal cations, while the variability of the products may be attributed to differences in the atomic radii of the Na⁺ (99 pm), K⁺ (138 pm). H⁺ is continuously consumed at the cathode resulting in an increase in pH. Ca²⁺ and Mg²⁺ are adsorbed on the cathode surface and form precipitates, leading to poisoning of the active site of the cathode. In fact, their inhibition principles were different. Ca²⁺ and Mg²⁺ at a concentration of 1 mmol/L were distributed into the ENRA system, and it was found that calcium ions had a great influence on the selectivity of NH₄⁺. The reason for this difference was that Ca²⁺ was mainly adsorbed to the catalyst surface to poison the active sites, but Mg²⁺ hindered ENRA by adhering to the cathode surface after being precipitated from the electrolyte [56].

Similarly, anions reduce the rate of nitrate reduction in the order I⁻ > Br⁻ > Cl⁻ > F⁻ (@ −1.8 V) [53]. Because some anions compete for adsorption sites or influence the magnitude of the electrostatic hindrance effect (characteristic enthalpy of adsorption of each anion on a particular metal cathode), which affects the electrochemical reduction of nitrate [57]. I⁻ would not form ion pairs with cations, Cl⁻ could reduce the competitive reaction. But SO₄²⁻ would compete with NO₃⁻ for the active center on the electrode surface in a certain potential range [54,57]. Among them, Cl⁻, which is present in large quantities in wastewater, also plays a distinctive role in electrochemical nitrate reduction. Easily, Cl⁻ reacts at the anode to produce HClO. It can be used not only for disinfection of water, but also for oxidation of NH₄⁺ to produce N₂ [53,58].

Nitrate-rich wastewater is widely distributed in ammunition industry/laboratories, pharmaceutical industry, mining sites, steel industry, electroplating and phosphorus industry, and the acidity and alkalinity of them vary [59]. For example, nitric acid which is commonly used in the electroplating industry to clean metals will produce a typically acidic wastewater, while wastewater from industries such as beverages, pulp, paper, and cement is mostly alkaline [60].

The electrochemical reduction of nitrate can occur over a wide pH ranging from highly acidic to highly basic, but its reduction kinetics and products vary at different pH [61–64]. Under acidic conditions, the HER reduces the Faradic efficiency of nitrate reduction because of competition for electrons [65,66]. Although the direct reaction of ENRA is affected by HER, the presence of indirect catalytic processes under acidic conditions can enhance the reduction kinetics [25]. However, the pH of the solution is not constant during the nitrate reduction process. The electrode interface and solution pH can even increase significantly as protons are consumed or hydroxides are produced [67–69]. During the process, the proton donor may change from H⁺ to H₂O, when the base-catalyzed reaction is more favorable (the acid-catalyzed reaction is inhibited) [32]. In addition, pH has a profound effect in terms of product. Studies have shown that under alkaline conditions, nitrate reduction products are dominated by NO₂⁻. The elevated proton concentration leads to the production of N₂, N₂O and NH₄⁺ [70]. Similarly, Dortsiou, Katsounaros, Polatides and Kyriacou [71] observed that the main products were nitrogen, nitrous oxide and nitrite at pH > 4. Whereas, when pH between 0 and 4, ammonia and hydroxylamine were the main products. However, drastic changes in the pH of the solution pose the risk of altered reaction mechanisms. Ideally, highly concentrated buffer solutions can be used to mitigate changes in interface and solution pH.

The stirring intensity of the reaction system also had a significant effect on the efficiency of nitrate reduction. The nitrate undergoes a reduction reaction on the cathode surface. The anion migrates towards the anode under the electric field, which is in the opposite direction to the mass transfer. Therefore, stirring can make the nitrate in the solution overcome the electric field repulsion and distribute it evenly, increasing the chance of contact between nitrate and cathode. At the same time, there is an equilibrium NH₄⁺ + H₂O ↔ NH₃ + H⁺ in the system [72], and vigorous stirring can help the reaction to occur continuously to the right. Reduction products from the cathode (e.g., H₂) cover the surface of the electrode after a period of time. Stirring encourages these bubbles to escape, increasing the contact area between the nitrate and electrode. However, too vigorous stirring may affect the stability of the electrodes. From the perspective of energy saving, continuing to increase the stirring rate will also increase the cost in practical engineering applications. Therefore, it is necessary to adjust the stirring intensity according to the reactor volume, which can make the solution fully mixed, but also consider energy saving.

In addition to the limitations of the reactor configuration, the electrode spacing also affects the nitrate reduction rate. Increasing the distance between electrodes will lead to an increase in transfer impedance, resulting in unnecessary economic losses [73]. However, when the distance between the pole plates is too small, the viscous effect of the solution between them is relatively enhanced. It causes a decrease in the turbulence of the liquid between the electrode plates, which leads to a decrease in the mass transfer capacity of nitrate or other nitrogen compounds between them. Therefore, the pole plate spacing can be adjusted according to the actual reactor and electrode area [74].

It is well known that the volume of the ENRA reactor varies from 30 mL to 500 mL at laboratory scale, which is far from sufficient for wastewater treatment. The main problems are the reactor configuration and electrode design if the scale of electrochemical devices is scaled up to hundreds of liters or even cubic. The operation of industrialized devices often adopts a continuous flow state, which requires the device to have excellent hydraulic flow state and less dead zone to achieve uniform reduction of nitrates. In terms of electrode design, reasonable electrode space utilization can also improve the turbulence of wastewater to promote mass transfer between nitrate and electrodes, improving nitrate reduction efficiency and ammonia generation efficiency. On the other hand, electrode materials need to be modified to ensure catalytic sites, which makes their preparation difficult and expensive. The electrode pieces are often present individually on a laboratory scale, so if they were to be scaled up directly, not only would there be technical difficulties with the modification, but the ohmic internal resistance of the electrodes would increase dramatically.

A pilot ENRA system with a working volume of 500 L was therefore constructed (Fig. 4a), the core solution was to place two electrode stacks with 13 electrode sheets near the bottom of the reactor to form an electrode module [16]. The form of stacked electrode module was used much earlier in bioelectrochemistry. The designed electrode module with a total volume of 1 m³ (L × W × H: 1.0 m × 1.0 m × 1.0 m) was constructed to contain eight pairs of sandwiched corrugated electrodes (cathode and anode of graphite coated stainless steel and carbon felt respectively), collector, power management system and unplasticised polyvinyl chloride cage (Fig. 4b) [81]. Similarly, our group has constructed an electrode module for pilot application (Fig. 4c), which consists of 10 sets of 1 m × 1 m carbon brush assemblies, with 5 sets of cathodes and 5 sets of anodes arranged crosswise with a controlled spacing of 10 cm. In summary, placing the electrode module assembly in the continuous flow wastewater treatment reactor can not only alleviate the difficulty of the batch reactor's limited processing capacity, but also solve the limitation of the number of reaction sites on the electrode surface. In addition, this electrode module alleviates the defect of unstable performance under the fluctuation of water quality and quantity due to its large space utilization rate. In terms of reactor configuration, perhaps strategies such as spiral tube reactors or multi-point feeds could improve the poor hydraulic flow regime.

Economic effect is an integral part of electrochemical reduction of nitrate to ammonia production. Mainstream ammonia synthesis processes are summarized in Table 2 [13,78,79,82–88]. The current mainstream ammonia synthesis process is HB, electrochemical nitrogen reduction reaction and ENRA. In general, HB process, as one of the greatest achievements in the 20^th century, cannot deny the economic effect it brings to human beings. But in terms of current national ideas and capital investment, the HB process is unsustainable. On the other hand, NRR is less efficient in overall ammonia production due to its poor nitrogen solubility and high inertness, not to mention the primary competition (HER) that easily occurs on most catalysts. The economic effect of ENRA is particularly significant. First, ENRA can be applied to the treatment of nitrate brine to realize the concept of turning waste into treasure. Secondly, nitrate is widely present in biological treatment tail water, which lays the foundation for its application. Third, the construction of the pilot ENRA system broke through the laboratory barriers. A more detailed economic assessment will be referred to below. Economic indicators involved in ENRA are electrical energy output, electrode cost and ammonia product return. It was found that the cost-effectiveness ratio of electric energy-driven ENRA in neutral medium was 0.45 (input to output ratio) without considering the cost of cathode materials (BCN@Ni). At this time, the faradaic efficiency of ammonia was only 59.2%, which makes it difficult for the subsequent extraction of ammonia products [76]. Another study, the cost-effectiveness ratio and faradaic efficiency of the electrical-driven ENRA were 0.498% and 82.68% without considering the cost of Pd₁₀Cu/BCN as cathode materials [13]. Therefore, it was understandable to sacrifice cost-effectiveness in order to obtain high value-added or high-yield ammonia. High value-added ammonia often exists in the form of fertilizers such as NH₄NO₃ and (NH₄)₂SO₄, Fig. 5a demonstrated the economic cost of electroreduction of nitrate to NH₄NO₃ based on a given price of electricity, cell efficiency, and total cell potential applied. The results show that it was feasible to obtain NH₄NO₃ with ENRA technology at low electricity costs and reasonable Faradaic efficiencies [30]. Fig. 5b showed the profitable area of power output and energy efficiency functions (energy-related parameters determined by NH₃ recovery/yield as a function of electricity cost) based on the ammonia recovery product as (NH₄)₂SO₄. Furthermore, evaluating its system economics in terms of energy-related parameters and corresponding electrical energy costs as a function of different nitrate concentrations, ENRA was found to be profitable only at high concentrations of nitrate (1 mol/L) (Fig. 5c) [75]. This series of economic indicators does not even take into account the cost of electrode materials, system maintenance costs, separation and concentration costs. It will undoubtedly bring predictable resistance to its large-scale application. Of course, there are also feasible methods: (1) Take the membrane concentrate as the treatment object. The membrane concentrate is the residue of the wastewater retained by the reverse osmosis or nanofiltration membranes, and the pollutants are mainly dissolved organic matter and a large number of inorganic ions. The low nitrate wastewater is also enriched with nitrate after retention, thus obtaining high nitrate wastewater. Of course, the concentration of other unhelpful pollutants will have adverse effects on the ENRA system, so when designing the treatment process, the arrangement of each treatment unit is particularly important; (2) Select the electrode materials that have been industrialized. In order to promote the large-scale application of ENRA, the economic benefits of the catalyst as the core of its reaction still need to be considered. Cu has been widely recognized as the catalyst with the highest activity for nitrate reduction and ammonia production [3], and the main component of brass mesh (5.7–22.9 ＄/m²) is copper (65%), which is undoubtedly a base material with great potential for industrialization. And the catalyst to be deposited can choose non-precious metal or non-metallic (Cu, Bi, Co, C, Ni, etc.) to further reduce the cost [22,89,90]. However, the economic effect of industrializing electrode materials or catalysts is not only reflected in the cost of the material, but also how to scale up and the long-term stability of the material. The expansion strategy has been mentioned in Section 6.1. The long-term stability of materials needs to be tested in actual engineering, carbon materials and BDD have application potential [91,92]. In practical applications, the replacement cycle, material cost, and comprehensive evaluation of ammonia production efficiency should be combined; (3) Optimize the system parameters to balance the power output and the value of the ammonia product (The details of the optimization parameters have been described in Sections 4 and 5).