Copper-Based Electrocatalysts for Nitrate Reduction to Ammonia

Ammonia (NH3) is a highly important industrial chemical used as fuel and fertilizer. The industrial synthesis of NH3 relies heavily on the Haber–Bosch route, which accounts for roughly 1.2% of global annual CO2 emissions. As an alternative route, the electrosynthesis of NH3 from nitrate anion (NO3−) reduction (NO3−RR) has drawn increasing attention, since NO3−RR from wastewater to produce NH3 can not only recycle waste into treasure but also alleviate the adverse effects of excessive NO3− contamination in the environment. This review presents contemporary views on the state of the art in electrocatalytic NO3− reduction over Cu-based nanostructured materials, discusses the merits of electrocatalytic performance, and summarizes current advances in the exploration of this technology using different strategies for nanostructured-material modification. The electrocatalytic mechanism of nitrate reduction is also reviewed here, especially with regard to copper-based catalysts.


Introduction
As an important industrial chemical, ammonia (NH 3 ) is widely used in synthetic drugs, fertilizers, dyes, plastics, and so on [1,2]. NH 3 is also regarded as a carbon-free and environmentally friendly clean-energy carrier of hydrogen, since NH 3 is easier to liquefy, store, and transport than hydrogen [3,4]. To date, the industrial synthesis of NH 3 has relied heavily on the Haber-Bosch route. This route requires harsh conditions of high temperature (400-600 • C) and high pressure (200-350 atm) [5]. In fact, the energy consumption of the Haber-Bosch process accounts for 1.4% of the world's total annual energy consumption, and the resulting carbon dioxide emissions are as high as 450 million tons, accounting for approximately 1.2% of global annual carbon dioxide emissions [6]. Therefore, researchers began to look for low-carbon and environmentally friendly alternatives for producing NH 3 that can be realized under normal temperature and pressure conditions. Among many alternatives, the method of producing NH 3 from nitrogen (N 2 ) and H 2 O driven by renewable electric energy (NRR) has gradually attracted the researchers' attention [7][8][9]. However, due to the high bond energy (941 kJ mol −1 ) of the N≡N bond and the low solubility of N 2 in water, the energy utilization of NRR in the aqueous environment is low, and its production yield of NH 3 is 2-3 orders of magnitude lower than the Haber-Bosch process. As the "Renewable Energy to Fuels through Utilization of Energy-dense Liquids" (REFUEL) program of the Department of Energy (DOE) demonstrates, the electrochemical synthesis of NH 3 requires the realization of an NH 3 -production Faraday efficiency exceeding 90%, an NH 3 yield of 10 −6 mol s −1 cm −2 , a current density of 300 mA cm −2 , and an energy efficiency of 60% for industrial production [10]. It is very difficult to achieve these indicators through the NRR process, which limits the further development of NRR [11]. For this reason, researchers have tried to apply other nitrogen-containing compounds as nitrogen sources to synthesize NH 3 [12]. Nitrate anion (NO 3 − ) is a good candidate, since the solubility of NO 3 − in water is high and the bond energy of N=O bond (204 kJ mol −1 ) is low [13]. The NO 3 − reduction process is a solid-liquid interface reaction, which is more conducive to mass transfer than a gas-liquid-solid interface reaction during NRR and has the potential to realize the target of the REFUEL program. NO 3 − is a pollutant that widely exists in industrial and agricultural wastewater and domestic sewage [14]. The residual NO 3 − in the environment will not only harm the water body ecosystem and lead to water eutrophication but will also pose a serious threat to human health [15]. Therefore, using NO 3 − as a nitrogen source for NH 3 synthesis is not only conducive to improving energy utilization efficiency and reducing greenhouse gas emissions; it can also alleviate the problem of NO 3 − pollution in the environment. In the natural environment, the conversion of NO 3 − to NH 3 can be biocatalyzed by a nitrate/nitrite reductase (such as Shewanella oneidensis cytochrome c nitrite reductase) under mild environmental conditions (a process known as bio-NRA) [16]. NH 3 can then be concentrated by physicochemical methods such as ion exchange adsorption and struvite precipitation and used for further industrial applications [6]. However, the biocatalytic process of bio-NRA takes a long time to effect the conversion, and the efficiency of NO 3 − removal is low. In addition, the complex structures of the bio-enzymes are difficult to synthesize artificially. Therefore, scientists began to simulate the function of microbial enzymes to design and develop artificially synthesized catalytic materials to satisfy the demands of industrial production [17]. Research on the electrocatalytic reduction of NO 3 − (NO 3 − RR) to NH 3 is also facing great challenges. NO 3 − RR involves the transfer of 8e − and 9 protons, during which various byproducts such as nitrite (NO 2 − ), nitrogen monoxide (NO), nitrogen (N 2 ), and nitrous oxide (N 2 O) are produced as well [18]. Meanwhile, there is competition between the hydrogen evolution reaction (HER) and NO 3 − RR. With the increase in the overpotential, HER will be increasingly strengthened, which further reduces the energy utilization efficiency of NH 3 formation. Therefore, it is of great significance to carry out in-depth research on the mechanism of electrocatalytic NO 3 − RR synthesis of NH 3 to rationally guide the development and design of catalysts with high catalytic activity and high NH 3 formation selectivity. 3 − RR N has a variety of stable forms of hydrides and oxides from −3 to +5 valence states. The thermodynamic relationship between these nitrides is shown in the Frost-Ebsworth diagram ( Figure 1) [13]. A series of reaction intermediate species and final products would be formed during NO 3 − reduction, and N 2 and NH 3 are the two most thermodynamically stable products, which can be obtained through the following Equations (1) and (2). Figure 1. Frost-Ebsworth diagram [19].

Reaction Mechanism of Electrocatalytic NO
According to the literature, NO3 − reduction to NO2 − is the rate-determining step of NO3 − RR (Figure 2) [20]. NO3 − in the solution is first adsorbed onto the cathode surface (Equation (3)), during which some disturbing anions in the solution may compete with NO3 − adsorption and inhibit NO3 − reduction [21]. The mass transfer rate of NO3 − from the solution to the electrode surface will also affect the adsorption process [22,23]. On the other hand, the number of active sites on the electrode's surface will affect the reaction rate [23]. Currently, researchers propose two reaction mechanisms for NO3 − RR to NH3: a direct reduction mechanism and an indirect autocatalytic reduction mechanism [19,24].

Direct Reduction Pathway
In the direct reduction pathway, the adsorbed NO3 − can be reduced by proton-coupling electrons or by adsorbed H atoms (*H) [19]. As shown in Equations (4)-(6), in the According to the literature, NO 3 − reduction to NO 2 − is the rate-determining step of NO 3 − RR (Figure 2) [20]. NO 3 − in the solution is first adsorbed onto the cathode surface (Equation (3)), during which some disturbing anions in the solution may compete with NO 3 − adsorption and inhibit NO 3 − reduction [21]. The mass transfer rate of NO 3 − from the solution to the electrode surface will also affect the adsorption process [22,23]. On the other hand, the number of active sites on the electrode's surface will affect the reaction rate [23]. Currently, researchers propose two reaction mechanisms for NO 3 − RR to NH 3 : a direct reduction mechanism and an indirect autocatalytic reduction mechanism [19,24]. According to the literature, NO3 − reduction to NO2 − is the rate-determining step o NO3 − RR ( Figure 2) [20]. NO3 − in the solution is first adsorbed onto the cathode surface (Equation (3)), during which some disturbing anions in the solution may compete with NO3 − adsorption and inhibit NO3 − reduction [21]. The mass transfer rate of NO3 − from the solution to the electrode surface will also affect the adsorption process [22,23]. On the other hand, the number of active sites on the electrode's surface will affect the reaction rate [23]. Currently, researchers propose two reaction mechanisms for NO3 − RR to NH3: a direct reduction mechanism and an indirect autocatalytic reduction mechanism [19,24].

Direct Reduction Pathway
In the direct reduction pathway, the adsorbed NO3 − can be reduced by proton-cou pling electrons or by adsorbed H atoms (*H) [19]. As shown in Equations (4)-(6), in the

Direct Reduction Pathway
In the direct reduction pathway, the adsorbed NO 3 − can be reduced by protoncoupling electrons or by adsorbed H atoms (*H) [19]. As shown in Equations (4)- (6), in the proton-coupling electron reduction pathway, the process of NO 3 − being reduced to NO 2 − involves three steps: (a) electrochemical (Equation (4)), (b) chemical (Equation (5)), and (c) electrochemical process (Equation (6)) [25][26][27]. Koper et al. [20] found that the Tafel slope of NO 3 − RR over transition metals in acidic solution is approximately 120 mV dec −1 , and then they speculated that the first electron-transfer step is the rate-determining step. The slow kinetics of this step was attributed to the high energy of the lowest unoccupied π* orbital (LUMO π*) of NO 3 − , which made it difficult for electron transfer from the electrode surface into the LUMO π* of *NO 3 − [28]. Cu with highly occupied d orbitals has a closed energy level with LUMO π* of NO 3 − . The electrons are easily transferred from the unclosed d orbital of Cu to the adsorbed NO 3 − , which is conducive to the electrochemical reduction of NO 3 − . NO 3 − adsorption: NO 2 − is a relatively stable intermediate which will become an unstable anion radical NO 2 2− through a direct electron transfer reaction pathway (Equation (7)) [29]. Similarly to NO 3 2− , NO 2 2− will be rapidly hydrolyzed into adsorbed NO ad through Equation (8) (k = 1.0 × 10 5 s −1 ) [30]. NO 2 − →NO: NO ad is the key intermediate species that determines the selectivity of the final products, N 2 or NH 3 . In the Vooys-Koper pathway, NO ad reacts with NO aq in solution to form a short-term dimer byproduct, dinitrogen dioxide (HN 2 O 2 ) (Equation (9)) [31], which is then reduced to dinitrogen oxide (N 2 O) by secondary electron transfer (Equation (10)) [23,32]. N 2 O can be further reduced to the unstable N 2 O − (Equation (11)) [33]. This species will be readily transformed into the final product N 2 (Equation (12)) [34]. NO→N 2 -Vooys-Koper pathway: Based on the NO 3 − RR study on the Pt (100) crystal plane, researchers proposed the Duca-Feliu-Koper pathway, through which NO 3 − will be electrochemically reduced to N 2 . In this pathway, stable NH 2ad will be generated by the reduction of NO ad (Equation (13)) and coexists with NO ad [35,36]. NO ad and NH 2ad will recombine according to the Langmuir-Hinshelwood reaction to generate the transient substance nitrosoamide (NONH 2 ) (Equation (14)), which will be rapidly decomposed to N 2 (Equation (15)).
Among the mechanism studies for NH 3 formation, the electrochemical-electrochemical (EE) reaction mechanism consisting of several sequential direct electron-transfer reactions is the main pathway [31]. NO successively undergoes reductive hydrogenation to several intermediates, forming nitroxyl (HNO) (Equation (16)) [37], H 2 NO (Equation (17)) [31], and hydroxylamine (H 2 NOH) (Equation (18)) [38]. Hydroxylamine will be protonated (Equation (19)) and then rapidly electrochemically reduced into NH 3 (Equation (20)) [39]. NO→NH 3 : In addition to the above direct reduction pathway, another hydrogenation pathway using adsorbed active hydrogen species (H ad ) as a reductant is considered to be an important pathway to promote the production of NH 3 [40,41]. In the potential region where hydrogen adsorption occurs on the Pt catalyst, the formation of NH 3 is highly promoted [42], since H ad is stable on the electrode surface in that potential window. H ad will efficiently deoxygenate NO 3 − and then hydrogenate the intermediates into NH 3 . A hydrogenation mechanism is postulated according to the theoretical calculation results (Equations (21)-(26)): [43] However, how H ad participates in the NO 3 − RR process is still not clear, and the key intermediates of the reaction have not been experimentally proved. The intrinsic mechanism of the hydrogenation pathway remains to be further explored.

Indirectly Autocatalytic Reduction Pathway
The indirectly autocatalytic reduction pathway occurs only in the strongly acidic solution of high NO 3 − concentrations (1-4 mol L −1 ) with the presence of HNO 2 [23] and mainly affects the step involving the reduction of HNO 3 to HNO 2 . In this pathway, NO 3 − itself is not the electroactive species and does not participate in the electron transfer process. NO 3 − and HNO 2 − will be transformed into NO 2 or NO + that is postulated as the electroactive species. With the accumulation of electroactive intermediates, the reaction will be highly enhanced. These two active substances can be generated via the Vetter pathway and the Schmid pathway, respectively (Equations (27)-(34)) [19]: Vetter pathway: Schmid pathway: The following pathway of HNO 2 to NH 3 is similar to the direct reduction pathway mentioned previously. The indirect autocatalytic reduction mechanism could effectively accelerate the reaction rate of NO 3 − RR, which plays an important role in accelerating the conversion rate and reducing the reaction cost.

Research Status of Electrocatalytic NO 3
− RR to NH 3 Production

Evaluation Criteria for NO 3 − RR Performance
Since nitrogen oxides and NH 3 widely exist in air and water, N species pollution during the tests must be strenuously avoided to guarantee the reproducibility of the research results. MacFarlane and Ib Chorkendorff et al. [44,45] proposed a set of testing procedures for NH 3 detection during NRR to ensure the accuracy of experimental results. The interference of N species pollution must be eliminated, and it must be determined that N in the formed NH 3 is derived from N 2 rather than from pollution from other N sources (catalysts, electrolytes, air, etc.) [11]. Similarly, the source of N in the formed NH 3 in the NO 3 − RR process should also be strictly verified to obtain the real experimental results. Therefore, we need to have a unified understanding of the performance evaluation criteria for NH 3 synthesis during the NO 3 − RR process. There are two main factors affecting the NO 3 − RR test results: (1) pollution from other N sources and (2) the accuracy of the NH 3 detection method. The pollution from other N sources mainly comes from air, reaction devices, electrolytes, and catalysts. Strict requirements on the cleanliness of the experimental equipment and accessories are essential to absolutely exclude these contaminations. Gases and chemicals used in experiments need to be at least analytically pure. The N element in the catalyst is another common contaminant affecting the experimental results. Many researchers who study single-atom catalyst systems should pay special attention to this if they use the MNC as the single-atom catalyst (M is the metal). It is necessary to verify whether the N content in the catalyst is consistent before and after the reaction. In addition, the 15 N isotope labeling experiment is also one of the methods of confirming that the 15 N in the product 15 NH 3 comes from 15 NO 3 − instead of pollution from other N sources. The Macfarlane and Zhang groups [46,47] proposed strict experimental methods for the 15 N isotope labeling experiment for NRR. Quantitative detection methods for NH 3 include UV-Vis spectroscopic/colorimetric methods (Nessler reagent and indophenol blue method) [47,48], the o-phthalaldehyde fluorescence method [49], ion chromatography (IC) [50], ion selective electrode (ISE) [51], hydrogen nuclear magnetic resonance spectroscopy ( 1 H NMR) [46,52], etc. The accuracy of these detection methods can be affected by environmental factors. For example, the pH will affect the complexation of the chromogenic agent with NH 3 , thereby affecting the coloration. The presence of other cations may affect the peak signal of NH 4 + when using ion chromatography detection. Two different detection methods are usually required for a simultaneous detection to ensure the accuracy of the experimental results. This involves using isotope-labeled 15 NO 3 − as the N source for the reaction and then using 1 H NMR to detect the 15 NH 3 product so that we can track the source of N, which is currently considered to be the safest measurement.
The production yield per unit area (mmol h −1 cm −2 ) and the production yield per unit mass (mmol h −1 g −2 ) are the key indicators in evaluating the NH 3 production rate. Faradaic efficiency (%) and partial current density (mA cm −2 ) are important indicators for the measurement of the energy utilization efficiency of the electrocatalytic process. In addition, the overpotential (V vs. RHE), stability, and NO 3 − removal rate (%) of the reaction are also important factors in evaluating the overall catalytic performances.

Electrocatalysts Designed for NO 3 − RR to NH 3
Noble metal catalysts have shown good electrocatalytic performances in many electrocatalytic reactions, such as hydrogen evolution, alcohol oxidation, nitrogen reduction, and carbon dioxide reduction reactions. In their study of electrocatalytic NO 3 − RR, DIMA et al. [20] analyzed the catalytic performance of the Pt group catalysts and found that the corresponding NO 3 − RR activity decreased following the sequence Rh > Ru > Ir > Pd > Pt (Figure 3a-e). The adsorption of NO 3 − over Pt and Pd is weak, while the adsorption of *H is too strong, making it difficult for NO 3 − RR to compete with HER. Rh, Ru, and Ir have a higher ability to adsorb NO 3 − and so those catalysts show better catalytic activity. Li et al. [53] reported that nitrate electroreduction overstrained Ru nanoclusters gave a high rate of NH 3 formation (5.56 mol g cat −1 h −1 ). The primary contributor to such performance is the *H, which are generated by suppressing H-H dimerization during the water splitting enabled by the tensile lattice strains. The *H expedited NO 3 − -to-NH 3 conversion by hydrogenating intermediates of the rate-limiting steps at lower kinetic barriers. The strained nanostructures can maintain nearly 100% NH 3 selectivity at >120 mA cm −2 current densities for 100 h. Feliu et al. [42] found that electrocatalytic nitrate reduction was hardly perceptible on Pt (111), while electrocatalytic activity was improved with an increase in the step density of the Pt electrode surface. Inactivation was observed on the step sites in the presence of adsorbed NO 3 − -derived reduced adsorbates, i.e., adsorbed NO. It was, therefore, concluded that the electrocatalytically active NO 3 − species did not adsorb on the (111) terraces but on the (111) monoatomic steps (as shown in Figure 3f,g). In addition, Han et al. [54] prepared Pd nanocrystalline with well-designed facets that could act as an efficient NO 3 − RR electrocatalyst for ambient NH 3 synthesis and found that Pd (111) exhibited excellent activity and selectivity in reducing NO 3 − to NH 3 with a Faradaic efficiency of 79.91% and an NH 3 production of 0. 55 , which was 1.4 times higher than Pd (100) and 1.9 times higher than Pd (110), respectively. However, although many studies have been conducted on noble metals and some interesting results have been obtained, more effort should be made to lower the cost of the fabrication of catalysts and apply them to the industrial production of NH 3 formation. Therefore, many researchers have turned their attention to the development and design of non-precious metal catalysts. Many transition metal catalysts exhibit a good electrocatalytic activity for NO3 − RR. Rossmeissl et. al. [55] studied the thermodynamics of various transition metal catalysts based on DFT calculations and found that Cu has a good adsorption capability for the intermediate species such as *NO3 − , *NO2 − , and *NO (Figure 4a-d), which is conducive to the continuous reduction and transformation of NO3 − . In addition, the adsorption of *H over Cu is weak, which hinders HER competition. Compared with other catalysts (as shown in Figure 4e), *N on Cu is more easily hydrogenated to NOH, which indicates that it has a better NH3 selectivity [12,56]. Jiao et al. [57] experimentally investigated various transition metals (Fe, Co, Ni, Cu) for the electrochemical reduction of NO and N2O to reveal the role of electrocatalysts in determining the product selectivity. Specifically, Cu was highly selective toward NH3 formation with >80% Faradaic efficiency in NO electroreduction. Furthermore, a high NO coverage facilitated the N-N coupling reaction. In acidic electrolytes, the formation of NH3 is greatly favored, whereas N2 production is suppressed. Many transition metal catalysts exhibit a good electrocatalytic activity for NO 3 − RR. Rossmeissl et. al. [55] studied the thermodynamics of various transition metal catalysts based on DFT calculations and found that Cu has a good adsorption capability for the intermediate species such as *NO 3 − , *NO 2 − , and *NO (Figure 4a-d), which is conducive to the continuous reduction and transformation of NO 3 − . In addition, the adsorption of *H over Cu is weak, which hinders HER competition. Compared with other catalysts (as shown in Figure 4e), *N on Cu is more easily hydrogenated to NOH, which indicates that it has a better NH 3 selectivity [12,56]. Jiao et al. [57] experimentally investigated various transition metals (Fe, Co, Ni, Cu) for the electrochemical reduction of NO and N 2 O to reveal the role of electrocatalysts in determining the product selectivity. Specifically, Cu was highly selective toward NH 3 formation with >80% Faradaic efficiency in NO electroreduction. Furthermore, a high NO coverage facilitated the N-N coupling reaction. In acidic electrolytes, the formation of NH 3 is greatly favored, whereas N 2 production is suppressed. corresponding to ΔE*H [55]. (d) A two-dimensional activity map for producing NH3, where all the reaction free energies are shown at 0 V vs. RHE [12]. (e) The logarithm of partial current density of NH3 for NORR on different metal electrodes at 0 V vs. RHE [56].

Modifications for Cu-Based Catalysts
Currently, the NH3 production rate over Cu-based catalysts is far lower than the traditional Haber-Bosch method (200 mmol gcat −1 h −1 ), which hinders its further development in industrial applications. Due to the complex reaction process of NO3 − RR to NH3 and the low kinetics of NO3 − RR, the current density of NO3 − reduction over Cu-based electrocatalysts is mostly below 100 mA cm −2 , especially in solutions with a low NO3 − concentration. Therefore, it is urgent to improve the intrinsic activity of Cu-based catalysts for NH3 formation. Research published in the literature describes the following methods of modification for Cu-based catalysts, which have been studied extensively; they are summarized below.

Size Modulation
The catalytic activity of the catalyst is highly affected by its particle size [58]. By optimizing the particle size of the catalyst, the catalytic activity of the catalyst can be significantly improved. As shown in Figure 5, from nanoparticles to nanoclusters and even single-atom catalysts, the density of active sites per unit mass increases, normally giving an increased catalytic activity [58]. Fu et al. [59] reported that the current density of NO3 − reduction over Cu nanoparticles is 13 times higher than that over Cu foil, and the Faraday efficiency of NH3 production has also increased from 30% over Cu foil to approximately 61% over Cu nanoparticles. In addition, the particle size of the catalyst will affect the interaction of reactants with the catalyst's surface [60]. Yang et al. [61] prepared a Cu singleatom catalyst (Cu SAC) for NO3 − RR during which the maximum NH3 yield reached 0.26 mmol cm −2 h −1 (12.5 mol gCu −1 h −1 ) and the Faraday efficiency of NH3 production reached 84.7%. According to in situ X-ray absorption spectroscopy coupled with advanced  [55]. (d) A two-dimensional activity map for producing NH 3 , where all the reaction free energies are shown at 0 V vs. RHE [12]. (e) The logarithm of partial current density of NH 3 for NORR on different metal electrodes at 0 V vs. RHE [56].

Modifications for Cu-Based Catalysts
Currently, the NH 3 production rate over Cu-based catalysts is far lower than the traditional Haber-Bosch method (200 mmol g cat −1 h −1 ), which hinders its further development in industrial applications. Due to the complex reaction process of NO 3 − RR to NH 3 and the low kinetics of NO 3 − RR, the current density of NO 3 − reduction over Cu-based electrocatalysts is mostly below 100 mA cm −2 , especially in solutions with a low NO 3 − concentration. Therefore, it is urgent to improve the intrinsic activity of Cu-based catalysts for NH 3 formation. Research published in the literature describes the following methods of modification for Cu-based catalysts, which have been studied extensively; they are summarized below.

Size Modulation
The catalytic activity of the catalyst is highly affected by its particle size [58]. By optimizing the particle size of the catalyst, the catalytic activity of the catalyst can be significantly improved. As shown in Figure 5, from nanoparticles to nanoclusters and even single-atom catalysts, the density of active sites per unit mass increases, normally giving an increased catalytic activity [58]. Fu et al. [59] reported that the current density of NO 3 − reduction over Cu nanoparticles is 13 times higher than that over Cu foil, and the Faraday efficiency of NH 3 production has also increased from 30% over Cu foil to approximately 61% over Cu nanoparticles. In addition, the particle size of the catalyst will affect the interaction of reactants with the catalyst's surface [60]. Yang et al. [61] prepared a Cu single-atom catalyst (Cu SAC) for NO 3 − RR during which the maximum NH 3 yield reached 0.26 mmol cm −2 h −1 (12.5 mol g Cu −1 h −1 ) and the Faraday efficiency of NH 3 production reached 84.7%. According to in situ X-ray absorption spectroscopy coupled with advanced electron microscopy technology, it was found that Cu SAC would be reconstructed into Cu nanoparticles (Cu 13 ) with a particle size of approximately 5 nm during NO 3 − RR. DFT calculation results also showed that there was a high thermodynamic energy barrier of *NO 2 reduction over Cu SAC, while the thermodynamic energy barrier of *NO 3 to NH 3 on a Cu 13 cluster presented a continuous downward trend. These results showed that it was the Cu nanoparticles that were the active sites and not the Cu single atom sites. electron microscopy technology, it was found that Cu SAC would be reconstructed into Cu nanoparticles (Cu13) with a particle size of approximately 5 nm during NO3 − RR. DFT calculation results also showed that there was a high thermodynamic energy barrier of *NO2 reduction over Cu SAC, while the thermodynamic energy barrier of *NO3 to NH3 on a Cu13 cluster presented a continuous downward trend. These results showed that it was the Cu nanoparticles that were the active sites and not the Cu single atom sites.

Crystalline Facet Engineering
Feliu and Koper [42,62] found that NO3 − RR on the Pt-based catalyst was highly affected by the surface crystalline facet. Recent NO3 − RR production also proved that NO3 − RR on the Cu-based catalyst was sensitive to the crystalline facet. Hu et al. [44] studied the NO3 − RR and HER processes on different basal facets of Cu by using theoretical calculations and found that the competition between NO3 − RR and HER was affected by the solution pH. Cu (100) showed a catalytic performance of NO3 − reduction to NH3 in a strong acidic environment that was higher than that of the other basal facet, while Cu (111) had a higher catalytic activity in a neutral or alkaline environment ( Figure 6). The difference in NO3 − RR activities between the different crystalline facets of Cu are attributed to the different atomic spacing and electronic state of the surface atom. The Cu-Cu bond length on Cu (111) was closer to the distance of two oxygen atoms in NO3 − , which is more conducive to the adsorption of NO3 − on Cu (111). In addition, in the theoretical calculation, the d-band center on the Cu (111) was higher than that of Cu (100), due to which there

Crystalline Facet Engineering
Feliu and Koper [42,62] found that NO 3 − RR on the Pt-based catalyst was highly affected by the surface crystalline facet. Recent NO 3 − RR production also proved that NO 3 − RR on the Cu-based catalyst was sensitive to the crystalline facet. Hu et al. [44] studied the NO 3 − RR and HER processes on different basal facets of Cu by using theoretical calculations and found that the competition between NO 3 − RR and HER was affected by the solution pH. Cu (100) showed a catalytic performance of NO 3 − reduction to NH 3 in a strong acidic environment that was higher than that of the other basal facet, while Cu (111) had a higher catalytic activity in a neutral or alkaline environment ( Figure 6). The difference in NO 3 − RR activities between the different crystalline facets of Cu are attributed to the different atomic spacing and electronic state of the surface atom. The Cu-Cu bond length on Cu (111) was closer to the distance of two oxygen atoms in NO 3 − , which is more conducive to the adsorption of NO 3 − on Cu (111). In addition, in the theoretical calculation, the d-band center on the Cu (111) was higher than that of Cu (100), due to which there was a stronger adsorption ability to NO 3 − on Cu (111). The electronic state of the catalyst surface affected its adsorption energy to NO 3 − and other intermediate species that affected the NO 3 − RR activity. At a potential of 0 V vs. RHE, *NO→*NOH was theoretically proved as the rate-limiting step. The low adsorption energy of *NOH can reduce the reaction energy barrier. Fu et al. [59] prepared the Cu catalysts with different basal facets and found that Cu (111) exhibited its highest catalytic activity in alkaline electrolytes. Koper et al. [63] studied the performance of the electrocatalytic NO 3 − over Cu single crystal electrode in different pH media. In acidic and alkaline media, the onset potential of Cu (111) was slightly more positive than Cu (100). However, the HER activity on Cu (111) was relatively higher than Cu (100), leading to better NO 3 − RR performances over the latter crystalline facet within a wide electrode potential window. Based on the advantages of electrochemically controllable technology in nanocrystals synthesis, our research group prepared Cu nanocrystals with different crystal plane, including high-index faceted Cu tetrahexahedron (THH Cu NCs, mainly composed of (211) and (311) crystal planes), low-index faceted Cu cube (100), and Cu octahedron (111) [64]. We found that, compared with Cu cube (100) and Cu octahedron (111), the THH Cu NCs with higher surface stepped sites had a higher reduction current density and a higher Faraday efficiency of NH 3 formation of 98.3% (Figure 7). This is consistent with the results on the Pt catalyst reported by Feliu et al. [42]. The surface stepped sites of the catalyst are the main active center. The higher the density of the steps, the better the NO 3 − RR activity.

Surface Defects Engineering
By constructing the defective structure and decreasing the coordination number of the active site on the catalysts surface, NO3 − adsorption and activation on the surface can be significantly enhanced and the intrinsic activity of the catalysts can then be improved. Hu et al. [65] studied the effect of defect exposure on the surface of Cu (100) on NO3 − reduction activity. They found that under alkaline conditions, ultra-thin two-dimensional CuO (111) nanoribbons reconstituted into Cu (100) nanoribbons (Cu-NBS -100) in the process of NO3 − electrochemical reduction, and, due to the high surface energy of ultra-thin nanoribbons, certain surface defects were formed in the process of structure reconstruction. Cu (100) -D1~7 models with different defects were constructed for DFT calculation. After comprehensive study, it was found that the d-band center of Cu (100)-D upshifted towards the Fermi level after the construction of surface defects on which the adsorption of intermediate species was enhanced. The Gibbs free energy of Cu (100)-D7 for NO3 − was significantly lower than that of Cu (100). The defective sites on Cu (100)-D7 had a strong adsorption effect on *H intermediates, which makes *H difficult to desorb from the Cu surface, thus inhibiting HER. Xu et al. [66] also proved that Cu nanosheets with rich surface defects had a higher catalytic activity than those with smooth surfaces. At the potential of −1.3 V vs. SCE, the production rate of NH3 reached 781.25 µg h −1 mg −1 , and the Faraday efficiency arrived at 81.99%. Surface defects were able to change charge distribution on the catalyst surface and effectively improved electron transfer between the catalyst and

Surface Defects Engineering
By constructing the defective structure and decreasing the coordination number of the active site on the catalysts surface, NO 3 − adsorption and activation on the surface can be significantly enhanced and the intrinsic activity of the catalysts can then be improved. Hu et al. [65] studied the effect of defect exposure on the surface of Cu (100) on NO 3 − reduction activity. They found that under alkaline conditions, ultra-thin two-dimensional CuO (111) nanoribbons reconstituted into Cu (100) nanoribbons (Cu-NBS -100) in the process of NO 3 − electrochemical reduction, and, due to the high surface energy of ultra-thin nanoribbons, certain surface defects were formed in the process of structure reconstruction. Cu (100) -D1~7 models with different defects were constructed for DFT calculation. After comprehensive study, it was found that the d-band center of Cu (100)-D upshifted towards the Fermi level after the construction of surface defects on which the adsorption of intermediate species was enhanced. The Gibbs free energy of Cu (100)-D7 for NO 3 − was significantly lower than that of Cu (100). The defective sites on Cu (100)-D7 had a strong adsorption effect on *H intermediates, which makes *H difficult to desorb from the Cu surface, thus inhibiting HER. Xu et al. [66] also proved that Cu nanosheets with rich surface defects had a higher catalytic activity than those with smooth surfaces. At the potential of −1.3 V vs. SCE, the production rate of NH 3 reached 781.25 µg h −1 mg −1 , and the Faraday efficiency arrived at 81.99%. Surface defects were able to change charge distribution on the catalyst surface and effectively improved electron transfer between the catalyst and the adsorbed species. They were able to enhance the adsorption and conversion of reaction intermediates on the catalyst surface [67]. In addition, surface atomic defects also brought more irregular edge steps to the catalyst surface and increased the density of surface-active sites.

Interfacial Engineering
In photocatalysis, constructing heterojunction structures to apply interfacial interactions to improve photocatalytic activity is a common catalyst-design strategy [68][69][70]. In electrocatalysis, the interfacial interaction between different phases also has a significant impact on catalytic performances [71,72]. Deng et al. [73] proved that the electrocatalytic NO 3 − RR activity of CoP can be effectively enhanced by constructing a p-n heterojunction of a CoP/TiO 2 nanoarray on titanium plate (CoP/TiO 2 @TP). According to DFT, the p-n heterojunction in CoP/TiO 2 could induce the charge redistribution of CoP and TiO 2 , enhance the electron transport between the interfaces, effectively reduce the free energy of each reaction step, and improve the performance of NO 3 − RR. Benefiting from the interaction of heterostructures, such CoP/TiO 2 @TP exhibited an excellent Faradaic efficiency of 95.0% and a large NH 3 yield as high as 499.8 µmol h −1 cm −2 , which was superior to CoP@TP and TiO 2 @TP. For a Cu-based catalyst, Liu et al. [74] grew Pd particles on CuO nano-olives (Pd/CuO NOs) to form a Pd/CuO NOs catalyst with a unique Pd/CuO heterogeneous interface for nitrite reduction (NO 2 − RR). The heterogeneous interface formation of Pd/CuO NOs was analyzed by means of XPS characterization and DFT calculation. In summary, the optimization of Cu catalysts by various modification methods can significantly improve the Faraday efficiency of NH 3 production, but it was also found that the current density of NO 3 − RR and the NH 3 production rate were still too low to be applied in industrial applications (Table 1). Moreover, the required overpotential was also too high and was not conducive to being powered by regeneration energy. To this end, researchers have gradually designed Cu-based bimetallic catalysts by introducing second metals, and tried to further improve the catalytic activity of Cu-based bimetallic catalysts by regulating their electronic structure and synergistic effect, and they have promoted the further development of NO 3 − RR.

Electronic Structure Modulation
Wang et al. [82] designed Cu 100−x Ni x alloy with different proportions of the Cu and Ni elements and investigated the adsorption energy of intermediate products over the Cu 100−x Ni x alloy. Since the NO 3 − RR activity on Ni is weak, the increase in the Ni element proportion should decrease the number of active sites on the surface, possibly resulting in a decreased performance. However, compared with pure Cu, Cu 50 Ni 50 showed an increase of 0.12 V in the half-wave potential and a sixfold increase in activity at 0 V vs. RHE. The electronic structure of Cu 100−x Ni x was analyzed by performing X-ray photoelectron spectroscopy (XPS), X-ray adsorption spectroscopy (XAS), and ultraviolet photoelectron spectroscopy (UPS). It was found that the binding energy of Cu 2p in Cu 100−x Ni x alloy shifted to a lower binding energy with the increase in the Ni proportion. The binding energy of the Ni 2p shifted towards a higher binding energy, indicating that the charge in the Cu 100−x Ni x alloy was redistributed and electrons were transferred from the Ni to the Cu. At the same time, the Cu d-band center upshifted, which enhanced the adsorption of the intermediate products on the Cu. Further study by means of DFT calculation found that the adsorption energy of the intermediate products showed a volcanic distribution with the overall reactivity. When the Ni accounted for 50%, the catalyst had moderate adsorption on all intermediate species, and then the highest reaction performances arrived. The establishment of the relationship between the electronic structure of the catalysts and NO 3 − RR activity provided an idea for guiding the design of novel NO 3 − RR catalysts. Ge et al. [83] reported polyallylamine (PA) functionalized RhCu bimetallic nanocubes (PA-RhCu cNCs) as a robust electrocatalyst for the reduction of NO 3 − to NH 3 . PA-RhCu cNCs showed a remarkable NH 3 production yield of 2.40 mg h −1 mg cat −1 and a high Faradaic efficiency of 93.7% at 0.05 V. DFT calculations and experimental results indicated that Cu and PA (adsorbed amino) co-regulated the Rh d-band center, which slightly weakens the adsorption energy of reaction-related species on Rh. These findings may open an avenue to construct other advanced catalysts based on organic molecule-mediated interfacial engineering. Goldsmith's group [84] predicted NO 3 − RR activity by calculating the adsorption energy of N and O atoms on transition metals. They proposed that alloyed Fe 3 Ru, Fe 3 Ni, Fe 3 Cu, and Pt 3 Ru have a strong adsorption capability to N and O atoms and suggested that those bimetals should be the potential NO 3 − reduction catalysts. Based on the above analysis, the appropriate adsorption energy of catalysts for intermediate species such as *NO 3 , *H, and *NH 3 is the key factor affecting their performance, which can be achieved by modulating the electronic structure of bimetal alloys to affect the d-band center or by forming a poor/rich-electron center.

Synergistic Effect
Protons (8e − ) are involved in the process of NO 3 − reduction to NH 3 , giving a series of elementary reactions before the final products are formed. The active sites on single metallic catalysts are not well adapted to simultaneous multiple elementary steps. To this end, the construction of a synergistic effect between multiple active sites become one of the research hotspots in the field of NO 3 − RR. Precious metals alloying with non-precious metals is one of the most useful strategies for fabricating bimetallic catalysts ( Table 2). The introduction of precious metals into Cu metal was proved to facilitate a much higher catalytic activity than that of pure Cu [85][86][87][88]. Kerkeni et al. [89] modified the platinum catalyst's surface through a deposition on the Cu (sub)monolayer of a certain amount of Cu (Cu ads /PT) using electrochemical methods. They found that the intrinsic catalytic activity (at t = 0) and the NH 3 selectivity of Cu ads /Pt depended on the fraction of the platinum surface occupied by copper adatoms (θ Cu ). The highest activities and selectivity of NH 3 were obtained with θ Cu ≈ 0.4-0.6. Moreover, higher activities were obtained in the reduction of NO 3 − than in the reduction of NO 2 − . Barrabés [85] and Wang et al. [90] believed that the existence of Cu was conducive to NO 3 − adsorption and reduction into NO 2 − and that NO 2 − would then be transferred to Pt/Pd to continue being hydrogenated. The presence of Pt/Pd would produce *H, which was conducive to maintaining the existence of Cu 0 activity and promoting the intermediate species to further hydrogenation to N 2 or NH 3 . Luo et al. [91] designed the Cu nanowire-supported Rh cluster and singleatom electrocatalyst (Rh@Cu) for a NO 3 − reduction reaction. Benefiting from the catalytic cooperation between Rh and Cu, whereby Rh activates the hydrogenation of the *NO intermediate on Cu, Rh@Cu systems achieved a Faradaic efficiency up to 93% at −0.2 V vs. an RHE and NH 3 yield rate of 1.27 mmol h −1 cm −2 . Detailed investigations by means of electron paramagnetic resonance, in situ infrared spectroscopy, differential electrochemical mass spectrometry, and DFT calculations suggested that the high activity originates from the synergistic cooperation between the Rh and Cu sites, whereby adsorbed hydrogen on the Rh site transferred to vicinal *NO intermediate species adsorbed on the Cu. Schuhmann et al. [92] presented a design concept of tandem catalysts, which involves coupling intermediate phases of different transition metals, existing at low applied overpotentials, as cooperative active sites that enable cascade NO 3 − -to-NH 3 conversion, in turn, avoiding the generally encountered scaling relations. They implemented the concept by means of the electrochemical transformation of Cu−Co binary sulfides into potentialdependent core−shell Cu/CuO x and Co/CoO phases. Electrochemical evaluation, kinetic studies, and in situ Raman spectra reveal that the inner Cu/CuO x phases preferentially catalyzed NO 3 − reduction to NO 2 − , which was rapidly reduced to NH 3 at the nearby Co/CoO shell. This unique tandem catalyst system led to a NO 3 − -to-NH 3 Faradaic efficiency of 93.3 ± 2.1% in a wide range of NO 3 − concentrations at pH 13, a high NH 3 yield rate of 1.17 mmol cm −2 h −1 in 0.1 M NO 3 − at −0.175 V vs. RHE, and a half-cell energy efficiency of~36%, surpassing most previous reports.

Study on Mechanism of Electrocatalysis of NO 3 − RR Synthesis of NH 3 by Cu-Based Catalyst
Hu et al. [44] studied the pathway of the electroreduction of NO 3 − to NH 3 by employing DFT calculation. According to their comprehensive thermodynamic and dynamic analysis, the pathway of NO 3 − RR on Cu should be as follows: NO 3 − → *NO 3 → *NO 2 → *NOH → *NHOH → *NH → *NH 2 → *NH 3 → NH 3 (g) (Figure 8). Butcher et al. [99] detected a series of intermediates in the NO 3 − RR process in the acid media by means of SHINERS analysis, including NO 2 − and HNO, NH 2 , NH 4 + , etc. This was consistent with DFT calculation results. In an alkaline medium, Koper et al. [63] observed the appearance of hydroxylamine on the Cu catalyst using Fourier infrared spectra analysis, indicating that hydroxylamine was an intermediate species in the alkaline environment.

Study on Mechanism of Electrocatalysis of NO3 − RR Synthesis of NH3 by Cu-Based Catalyst
Hu et al. [44] studied the pathway of the electroreduction of NO3 − to NH3 by employing DFT calculation. According to their comprehensive thermodynamic and dynamic analysis, the pathway of NO3 − RR on Cu should be as follows: NO3 − →*NO3 →*NO2 →*NOH →*NHOH →*NH →*NH2 →* NH3 → NH3 (g) (Figure 8). Butcher et al. [99] detected a series of intermediates in the NO3 − RR process in the acid media by means of SHINERS analysis, including NO2 − and HNO, NH2, NH4 + , etc. This was consistent with DFT calculation results. In an alkaline medium, Koper et al. [63] observed the appearance of hydroxylamine on the Cu catalyst using Fourier infrared spectra analysis, indicating that hydroxylamine was an intermediate species in the alkaline environment. Gewirth et al. [100] analyzed the structural evolution of the adsorption intermediates on the surface of the Cu (100) by using an in situ electrochemical scanning tunnel microscope (EC-STM). They intuitively observed the process of adsorption of NO 3 − being transformed into NO 2 − . The adsorption of oxyanions was always accompanied by the synergistic adsorption of H 2 O or H 3 O + [101]. H 2 O or H 3 O + can stabilize the structure of oxyanions by forming hydrogen bonds with the lone electron pairs of oxygen atoms [102]. Researchers proposed that the NO 3 − adsorption layer on Cu (100) was also involved with H 3 O + . In addition, DFT calculation results showed that NO 3 − and NO 2 − were adsorbed on the Cu (100) surface by the bridging of their two oxygen atoms. Koper et al. [20] conducted an electrochemically dynamic analysis of the NO 3 − RR process using cyclic voltammetry and found that the Tafel slope on the Cu-based catalyst was slightly higher than the 120 mV dec −1 . Therefore, they proposed that the first electron transfer step was the rate-limiting step. There are two possible mechanisms: *H reduction (Equations (35) and (36)) and the proton-coupling electron reduction mechanism (Equations (37) and (38)). *H reduction: Proton-coupling electron reduction: HNO − 3 ad → adsorbed intermedia (38) The HER performance over the Cu catalyst is relatively poor, so it is difficult to produce *H species. Therefore, it is generally believed that the proton-coupling electron reduction mechanism is most effective on a pure Cu catalyst. For catalysts that are more susceptible to activating protons, such as the precious metals contained in bimetals, *H was more conducive to NO 3 − reduction. The existence of other anions/cations in electrolytes usually affects the target catalytic reactions. For example, the existence of Cl − ions would lead to different final products. Butcher et al. [99], analyzing to Raman spectra, found that there were more bands of N-H x species on the Cu basal facets after adding 10 mmol L −1 Cl − , revealing that the presence of Cl − was beneficial for NH 3 formation. Gao et al. [103] found that the amount of NH 3 formation was dramatically decreased, although a similar removal efficiency of NO 3 − was obtained when 100 mg L −1 Cl − was added. As the Cl − ion concentration increased, the NH 3 selectivity rapidly decreased. This was due to the Cl − indirectly oxidizing NH 3 into N 2 . The Cl − was oxidized by the anode to produce Cl 2 (Equation (39)), which was then transformed into HOCl with strong oxidation (Equation (40)) [104]. The HOCl could efficiently oxidize the NH 3 into N 2 (Equation (41)).
Currently, mechanistic studies of NO 3 − RR over Cu-based catalysts are relatively scarce, and most of them have followed the results from Pt-based catalysts. Therefore, the mechanism of electrocatalytic NO 3 − reduction to NH 3 over Cu-based catalysts requires further study.

Conclusions and Perspectives
Electrocatalytic NO 3 − RR is a clean and pollution-free process for NH 3 production, which has the potential to realize large-scale industrial NH 3 production and could solve the problem of NO 3 − pollution at the same time. This review mainly presents contemporary views on the state of the art in electrocatalytic NO 3 − reduction over Cu-based nanostructured materials. The effective methods, which were reported to modify the performance of Cu-based catalysts for electrocatalytic NO 3 − RR to NH 3 , include size modulation, crystalline facet engineering, surface defect engineering, interfacial engineering, electronic structure modulation, and synergistic effect. A reduction in metal catalyst particle size can effectively improve the utilization efficiency of atoms and enhance reactivity. Theoretically, the single-atom catalyst has nearly 100% atomic utilization efficiency. However, for Cu single-atom catalysts, the active sites were Cu nanoparticles instead of the Cu single atom sites. For the optimization of crystal surfaces, Cu nanocrystalline catalysts with a high crystal index have higher activity than those with a low crystal index. Cu catalysts with a high crystal index can be constructed to further improve the performance of NH 3 synthesis. Moreover, by constructing the defective structure and decreasing the coordination number of the active site on the catalyst's surface, NO 3 − adsorption and activation on the surface can be significantly enhanced, and the intrinsic activity of the catalysts can then be improved. Cu active sites can be optimized effectively by the three methods mentioned above. However, since the reduction of NO 3 − to NH 3 involves the transfer of nine protons and eight electrons, it is difficult for the single active site to work effectively with simultaneous multiple elementary steps. To this end, catalysts with multiple active sites were designed. The heterogeneous interface between the Cu-based catalysts and other materials could promote local charge polarization by means of the formation of a built-in electric field and change the adsorption of intermediate species of NO 3 − RR. The alloying of Cu with a secondary metal can regulate the electronic structure of Cu and the synergistic effect. Additionally, different elements/active sites with different activities may have synergistic effects by providing a reduction in intermediate species or tandem reaction mechanisms to effectively reduce the reaction energy barrier. In conclusion, there are broad development spaces in the design and development of catalysts for the electrocatalytic NO 3 − RR of synthetic NH 3 , and this paper can provide a reference for subsequent research.
The research of nitrate reduction reaction has developed vigorously and made some progress in recent years, but there are still many problems. Below, we list some of the shortcomings and further research directions.
Research on the mechanism of NO 3 − RR is still insufficient. The reaction pathway of NO 3 − RR is affected by many environmental parameters, such as other ions, pH, NO 3 − concentration, and applied potential. Different systems accrue different basic pathways, which require a more systematic and detailed analysis, including the identification of key reaction intermediates by means of various in situ spectra and the postulation of basic reaction steps using theoretical calculations. Electrocatalysts that are more economical, efficient, and stable need to be developed. At present, although modified Cu-based catalysts show improved electrocatalytic reduction activities, there are oxidation, dissolution, and leaching problems in catalysts, and these lead to the decay of electrochemical activity and could cause adverse effects on the environment. In addition, structural evolution during electrocatalysis makes it more difficult to identify the real active sites and design strategies for protecting these active sites. Various electrochemical in situ spectral analyses and electron microscopic observations can help track the evolution of active sites during electrocatalysis and provide guidance and assistance for catalyst design. Furthermore, the predictive power of machine learning based on the analysis of large experimental data will help to develop the rapid screening of suitable and efficient NO 3 − RR catalysts. There are differences between the actual sewage and the experimental test electrolyte. The potential and limitations of electrocatalytic NO 3 − RR in practical water-remediation systems should be investigated by fully considering the effects of possible ion pollution, organic compound pollution, and solid sediment on NO 3 − RR in actual sewage and then designing the corresponding solutions. For example, pretreatment devices could be designed to purify the pollution.
It is necessary to design and build a reasonable and efficient reaction system and device. First, in order to form a complete chemical reaction process with NO 3 − RR, another set of electrocatalytic reactions (such as oxygen evolution) would be required, which would not only increase the complexity and cost of the reaction but also mean that the oxidation process and reaction byproducts might interfere with NO 3 − reduction. Therefore, we need to optimize the electrolytic cell. In addition, it is necessary to overcome the influence of mass transfer on scale-up experiments. The design of more efficient flow electrolytic cells is also one of the keys to improving the efficiency of NO 3 − RR. The design and evaluation of the whole chain of the NH 3 synthesis process by electrochemical NO 3 − RR require further work. At present, the low efficiency of NH 3 extraction is an important factor hindering the development of NH 3 production by NO 3 − RR. Therefore, the overall efficiency and the feasibility of NO 3 − RR in practical applications still need to be further evaluated. For example, the efficiency and cost of electrocatalytic NO 3 − RR technology and the Haber-Bosch process, as well as other important parameters related to scalability and commercial feasibility, should be investigated to make a comprehensive technological and economic evaluation.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Conflicts of Interest:
The authors declare no conflict of interest.