Electrocatalytic Nitrate Reduction for Brackish Groundwater Treatment: From Engineering Aspects to Implementation

Abstract

In recent years, nitrate has emerged as a significant groundwater pollutant due to its potential ecotoxicity. In particular, nitrate contamination of brackish groundwater poses a serious threat to both ecosystems and human health and remains difficult to treat. A promising, sustainable, and environmentally friendly solution when biological treatments are not applicable is the conversion of nitrate to harmless nitrogen (N₂) or ammonia (NH₃) as a nutrient by electrocatalytic nitrate reduction (eNO₃R) using solar photovoltaic energy. This review provides a comprehensive overview of the current advances in eNO₃R for the production of nitrogen and ammonia. The discussion begins with fundamental concepts, including a detailed examination of the mechanisms and pathways involved, supported by Density Functional Theory (DFT) to elucidate specific aspects of ammonium and nitrogen formation during the process. Furthermore, the integration of artificial intelligence (AI) and machine learning (ML) offers promising advancements in enhancing the predictive power of DFT, accelerating the discovery and optimization of novel catalysts. In this review, we also explore various electrode preparation methods and emphasize the importance of in situ characterization techniques to investigate surface phenomena during the reaction process. The review highlights numerous examples of copper-based catalysts and analyses their feasibility and effectiveness in ammonia production. It also explores strategies for the conversion of nitrate to N₂, focusing on nanoscale zerovalent iron as a selective material and the subsequent oxidation of the produced ammonia. Finally, this review addresses the implementation of the eNO₃R process for the treatment of brackish groundwater, discussing various challenges and providing reasonable opinions on how to overcome these obstacles. By synthesizing current research and practical examples, this review highlights the potential of eNO₃R as a viable solution to mitigate nitrate pollution and improve water quality.

1. Introduction

The water crisis is considered to be the most critical problem that requires attention and practical solutions, as more than one billion people do not have access to drinking water [1,2]. The reasons for the limited access to drinking water are usually the unequal distribution of water on Earth, with only 2.5% freshwater in the total water, and water pollution by dissolved organic and mineral substances [3]. Groundwater is typically regarded as one of the purest sources of drinking water in many countries, such as India [4], and the Anti-Atlas in Morocco [5]. However, these water resources are facing serious problems due to rapid urbanization and industrialization. One of these issues is nitrate contamination, which presents a serious threat to ecosystems and human health. Broadly speaking, nitrate accumulation in groundwater is caused by various anthropogenic and geogenic activities [6]. In fact, it is commonly agreed that agricultural activities are the most important source of nitrate pollution in groundwater [7,8,9,10]. Many regions of the world report a significant percentage of nitrates in groundwater that exceeds World Health Organization (WHO) standards. Hence, anthropogenic nitrate is an environmental handicap with serious consequences when it reaches the human body through water ingestion [11], especially a disease that occurs in the bloodstream of babies under 6 months of age, called infant cyanosis, methemoglobinemia, or blue babies [12]. The conversion process of nitrates to nitrites and then to nitrosamines in the human body leads to harmful effects, such as gastric cancer [13]. That is why the WHO considers drinking water safe in terms of nitrates when it has a quality ratio less than or equal to one ($\mathrm{QR}={\displaystyle \frac{{\mathrm{C}}_{\mathrm{nitrate}}}{50}}\le 1)$ [14].

To meet the permissible standard of nitrate concentration in drinking water, many physicochemical techniques have been adopted to remove nitrates from the water, such as adsorption [15,16], ion exchange [17,18], biological denitrification [19,20], anaerobic ammonium oxidation (Anammox) [21,22], chemical denitrification [23,24], catalytic denitrification [25,26,27], nanofiltration, and reverse osmosis [28,29]. Nevertheless, these processes suffer from serious limitations, including high installation and maintenance costs, biofouling, brine generation, membrane sensitivity, and the need for post-treatment [30]. Currently, eNO₃R can be perceived as the best way to treat contaminated water containing nitrogen compounds owing to their advantages, such as easy operation, remote control, minimal generation of secondary waste generation, and no potential for failure associated with high salinity, and these treatments can also be driven by renewable energy such as solar energy, hydropower, and wind power [31,32]. On the other hand, some disadvantages can minimize their effectiveness, such as low efficiency for higher nitrate contents and very long reaction time [33,34]. Numerous research investigations have centered their attention on the electrochemical synthesis of ammonia (${\mathrm{NH}}_3$) through the reduction in nitrate (${\mathrm{NO}}_3^{-}$) to ammonium (${\mathrm{NH}}_4^{+}$) in nitrate-contaminated water. This focus emerges as a promising avenue to alleviate the substantial energy demands entailed by the conventional Haber–Bosch industrial process, which produces ${\mathrm{NH}}_3$ from nitrogen (${\mathrm{N}}_2$) and hydrogen (${\mathrm{H}}_2$) [35]. Simultaneously, it offers a sustainable means of nitrogen recovery. The second strategy is eliminating ${\mathrm{NO}}_3^{-}$ from groundwater while concurrently converting it into nitrogen gas. This pursuit assumes significance due to its pivotal role in maintaining equilibrium within the nitrogen cycle (N-cycle). From a thermodynamic standpoint, ${\mathrm{N}}_2$ boasts the lowest solubility in water, simplifying its separation from aqueous mediums. Consequently, the electroreduction in ${\mathrm{NO}}_3^{-}$ into ${\mathrm{N}}_2$ emerges as a highly desirable approach, representing an efficient pathway for the generation of potable water and the restoration of equilibrium within the nitrogen cycle [36]. Thus, the present review focuses on introducing the mechanism of eNO₃R and providing a basic understanding of the complex multi-electron transfer required to successfully carry out the process, by discussing the trends in the catalytic activity of the eNO₃R to ammonia and nitrogen using Density Functional Theory (DFT). In addition, we explore the emerging role of machine learning (ML) and artificial intelligence (AI) in enhancing DFT methodologies, highlighting their potential to accelerate the discovery and optimization of efficient catalysts for nitrate reduction. As well, we review the various preparation methods for electrocatalysts and discuss advanced characterization techniques, with particular emphasis on in situ methods, to better understand catalyst behavior under reaction conditions. Furthermore, the most recent advancements of electrocatalysts for ${\mathrm{NH}}_3$ production by copper-based catalysts as a case study have been summarized, and the strategies to convert nitrate to nitrogen by selective materials as the case of nanoscale zero valent iron and the oxidation of ammonium generated during electrolysis are highlighted. Additionally, parameters that are influencing the reactivity and selectivity such as pH, coexistence ions, current density, and applied voltage are presented. Finally, the application of the process with the perspective, challenges, and technoeconomic analysis for brackish groundwater is analyzed.

As a result, the objective of this review is to propose a comprehensive overview of the current advances in eNO₃R for the production of nitrogen and ammonia. This includes a discussion of fundamental concepts, a detailed examination of the mechanisms and the potential catalysts, and an analysis of the pathways of ammonium and nitrogen formation with a focus on strategies for the conversion of nitrate to N₂. Finally, this review addresses the implementation of the eNO₃R process for the treatment of brackish groundwater as a typical example. By synthesizing current research and practical examples, this review highlights the potential of eNO₃R as a sustainable solution to mitigate nitrate pollution and improve water quality.

2. Reaction Mechanisms and Kinetics of Electrocatalytic Nitrate Reduction

The electrochemical nitrate reduction mechanism is a complicated multi-electron-transfer process, with 5-electron to form ${\mathrm{N}}_2$ and 8-electron to generate ${\mathrm{NH}}_4^{+}$ which involves a variety of reaction intermediates and products. According to the relevant standard redox potential of nitrate reduction and the respective possible products (

Table 1. Standard redox potentials for different ${\mathrm{NO}}_3^{-}$ reduction reactions.

Reduction Reaction	Standard Redox Potential
2${\mathrm{NO}}_3^{-}$(aq) + 12 H+ + 10 e− ⇌ ${\mathrm{N}}_2$(g) + 6 H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{N}}_2}^0$ = + 1.25 V
${\mathrm{NO}}_3^{-}$(aq) + 10 H+ + 8 e− ⇌ ${\mathrm{NH}}_4^{+}$(aq) + 3 H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{NH}}_4^{+}}^0$ = + 1.20 V
2${\mathrm{NO}}_3^{-}$(aq) + 10 H+ + 8 e− ⇌ ${\mathrm{N}}_2$O(g) + 5 H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{N}}_2\mathrm{O}}^0$ = + 1.12 V
${\mathrm{NO}}_3^{-}$(aq) + 4 H+ + 3 e− ⇌ NO(g) + 2 H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/\mathrm{NO}}^0$ = + 0.96 V
$2{ \mathrm{NO}}_3^{-}$(aq) + 10 H+ + 8 e− ⇌ ${\mathrm{H}}_2{\mathrm{N}}_2{\mathrm{O}}_2$ (g) + 4 H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{H}}_2{\mathrm{N}}_2{\mathrm{O}}_2}^0$ = + 0.90 V
${\mathrm{NO}}_3^{-}$(aq) + 2 H+ + 2 e− ⇌ ${\mathrm{NO}}_2^{-}$(aq) + H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{NO}}_2^{-}}^0$ = + 0.89 V
${\mathrm{NO}}_3^{-}$(aq)− + 9 H+ + 8e− ⇌ ${\mathrm{NH}}_3$(aq) + 3H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{NH}}_3}^0$ = + 0.82 V
${\mathrm{NO}}_3^{-}$(aq) + 2 H+ + e− ⇌ ${\mathrm{NO}}_2$(g) + H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{NO}}_2}^0$ = + 0.80 V
${\mathrm{NO}}_3^{-}$(aq) + 7 H+ + 6 e− ⇌ ${\mathrm{NH}}_2\mathrm{OH}$(aq) + 2 H2O(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{NH}}_2\mathrm{OH}}^0$ = + 0.67 V
${\mathrm{NO}}_3^{-}$(aq) + 7 H2O + 8 e− ⇌ ${\mathrm{NH}}_4\mathrm{OH}$(aq) + 9${\mathrm{OH}}^{-}$(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{NH}}_4\mathrm{OH} }^0$ = −0.12 V
$2{ \mathrm{NO}}_3^{-}$(aq) + 10 H2O + 14 e− ⇌ ${\mathrm{N}}_2{\mathrm{H}}_4$(aq) + 16${\mathrm{OH}}^{-}$(l)	${\mathrm{E}}_{{\mathrm{NO}}_3^{-}/{\mathrm{NH}}_4\mathrm{OH} }^0$ = −0.15 V

), the most stable compounds are ${\mathrm{N}}_2$ and ${\mathrm{NH}}_4^{+}$ which shows the highest equilibrium potential among all the possible products of ${\mathrm{NO}}_3^{-}$ reduction [37].

2.1. Mechanistic Insight into the Electrochemical Nitrate Reduction

From an electrochemical standpoint, several investigations have confirmed that eNO₃R can involve different pathways, which produce a variety of products that enclose nitrogen gas and ammonia as well as undesired species such as nitrite, hydrazine, hydroxylamine, nitric oxide, nitrous oxide, and others as depicted in Figure 1.

The electroreduction process of nitrates, as shown in various studies, initiates with the conversion of ${\mathrm{NO}}_3^{-}$ to ${\mathrm{NO}}_2^{-}$, which is commonly referred to as the rate-determining step (RDS). This step starts with the adsorption of nitrate as the electron is completely donated to the electrocatalyst surface. The adsorbed nitrate (${\mathrm{NO}}_3^{–}$_ads) could be reduced to the unstable nitrate dianion radical (^•${\mathrm{NO}}_3^{2–}$), which is hydrolyzed to form a nitrogen dioxide radical (${\mathrm{NO}}_2^{•}$) that is finally reduced to adsorbed nitrite (${\mathrm{NO}}_{2 \mathrm{ads}}^{–})$ playing a critical role in the overall reaction pathway as given in Equations (1)–(4) [38].

$${\mathrm{NO}}_{3 \mathrm{aq}}^{-}+ * \rightleftharpoons {\mathrm{NO}}_{3 \mathrm{ads} }^{-}$$

(1)

$${\mathrm{NO}}_{3 \mathrm{ads} }^{-}+{\mathrm{e}}^{-}\to {\mathrm{NO}}_{3 \mathrm{ads} }^{2-}$$

(2)

$${\mathrm{NO}}_{3 \mathrm{ads} }^{2-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{NO}}_{2 \mathrm{ads} }^{•}+2{ \mathrm{OH}}^{-}$$

(3)

$${\mathrm{NO}}_{2 \mathrm{ads} }^{•}+{\mathrm{e}}^{-}\to {\mathrm{NO}}_{2 \mathrm{ads} }^{-}+{\mathrm{H}}_2\mathrm{O}$$

(4)

The next step is the reduction in intermediate ${\mathrm{NO}}_{2 \mathrm{ads}}^{-}$ into adsorbed nitric oxide (${\mathrm{NO}}_{\mathrm{ads}}$) that is considered an important intermediate that opens an alternative reaction into three pathways (Figure 2) [39,40].

Duca et al. [41] indicated that the hydrogenation of ${\mathrm{NO}}_{\mathrm{ads}}$ to ${\mathrm{NH}}_{2 \mathrm{ads}}$, then ${\mathrm{NH}}_{2 \mathrm{ads}}$ reacts with ${\mathrm{NO}}_{\mathrm{ads}}$ and desorbs to form ${\mathrm{N}}_2$ (Equations (5) and (6)).

$${\mathrm{NO} }_{\mathrm{ads}}+4{\mathrm{e}}^{-}+3{ \mathrm{H}}_2\mathrm{O} \to {\mathrm{NH}}_{2, \mathrm{ads}}+4{ \mathrm{OH}}^{-}$$

(5)

$${\mathrm{NO} }_{\mathrm{ads}}+{\mathrm{NH}}_{2, \mathrm{ads}}\to {\mathrm{N}}_2+{ \mathrm{H}}_2\mathrm{O}$$

(6)

Furthermore, De Vooys et al. [40] showed that the mechanism proceeds by dimerization of ${\mathrm{NO}}_{\mathrm{ads}}$ to form ${\mathrm{N}}_2{\mathrm{O}}_2$, followed by protonation to ${\mathrm{HN}}_2{\mathrm{O}}_2$ which is then reduced electrochemically to ${\mathrm{N}}_2{\mathrm{O}}_{ \mathrm{ads}}$ as the following equations (Equations (7) and (8)).

$${\mathrm{NO} }_{\mathrm{ads}}+{\mathrm{NO} }_{\left(\mathrm{aq}\right)}+{ \mathrm{H}}^{+}+{ \mathrm{e}}^{-} \to {\mathrm{HN}}_2{\mathrm{O}}_{2, \mathrm{ads}}$$

(7)

$${\mathrm{HN}}_2{\mathrm{O}}_{2, \mathrm{ads}}+{ \mathrm{H}}^{+}+{ \mathrm{e}}^{-} \to {\mathrm{N}}_2{\mathrm{O}}_{ \mathrm{aq}}+{ \mathrm{H}}_2\mathrm{O}$$

(8)

Then, Chumanov et al. [42], confirmed that ${\mathrm{N}}_2{\mathrm{O}}_{ \mathrm{aq}}$ can then react with the hydrated electron and undergo the following reactions to form ${\mathrm{N}}_2$ (Equations (9)–(11)).

$${\mathrm{N}}_2{\mathrm{O}}_{ \mathrm{aq}}+{\mathrm{e}}^{-}\to {\mathrm{N}}_2{\mathrm{O}}^{-}$$

(9)

$${\mathrm{N}}_2{\mathrm{O}}^{-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{N}}_2+{\mathrm{OH}}^{-}+{\mathrm{OH}}^{•}$$

(10)

$${\mathrm{OH}}^{•}+{\mathrm{e}}^{-}\to {\mathrm{OH}}^{-}$$

(11)

The third proposed pathway is the formation of hydroxylamine (${\mathrm{NH}}_2\mathrm{OH}$) and ${\mathrm{NH}}_4^{+}$ due to the prevention of desorption of ${\mathrm{NO}}_{\mathrm{ads}}$ which allows its hydrogenation to $\mathrm{HNO}$, ${\mathrm{H}}_2\mathrm{NO}$, ${\mathrm{NH}}_2\mathrm{OH}$, and finally ${\mathrm{NH}}_4^{+}$ (Equations (12)–(16)) [40,43,44,45,46,47].

$${\mathrm{NO} }_{\mathrm{ads}}+{ \mathrm{H}}^{+}+{\mathrm{e}}^{-}\rightleftharpoons {\mathrm{HNO} }_{\mathrm{ads}}$$

(12)

$${\mathrm{HNO} }_{\mathrm{ads}}+{ \mathrm{H}}^{+}+{\mathrm{e}}^{-}\to {\mathrm{NH}}_2{\mathrm{OH} }_{\mathrm{ads}}$$

(13)

$${\mathrm{NH}}_2\mathrm{OH}+{\mathrm{H}}^{+}\rightleftharpoons {\mathrm{H}}_3{\mathrm{NOH}}^{+}$$

(14)

$${\mathrm{NH}}_2\mathrm{OH}+2{ \mathrm{H}}^{+}+{\mathrm{e}}^{-}\to {\mathrm{NH}}_3+{ \mathrm{H}}_2\mathrm{O}$$

(15)

$${\mathrm{NH}}_3+{\mathrm{H}}^{+}\rightleftharpoons {\mathrm{NH}}_4^{+}$$

(16)

On the other hand, Vetter and Schmid proposed two autocatalytic mechanisms (Equations (17)–(24)) due to the protonation of the unstable ${\mathrm{NO}}_{2 \mathrm{ads} }^{-}$ to nitrous acid in highly acidic media, which creates a sufficient amount of ${\mathrm{HNO}}_2$ that dominates over the ${\mathrm{NO}}_{2 }^{-}$, leading to the spontaneous formation of several products, including nitric oxide ($\mathrm{NO}$), nitrogen dioxide (${\mathrm{NO}}_2$), and ${\mathrm{HNO}}_2$ as a result of disproportionation reactions [48].

Vetter mechanism

$${\mathrm{NO}}_2+{\mathrm{e}}^{-}\to {\mathrm{NO}}_2^{-}$$

(17)

$${\mathrm{NO}}_2^{-}+{\mathrm{H}}^{+}\rightleftharpoons {\mathrm{HNO}}_2$$

(18)

$${\mathrm{HNO}}_2+{\mathrm{HNO}}_3\to {\mathrm{N}}_2{\mathrm{O}}_4+{ \mathrm{H}}_2\mathrm{O}$$

(19)

$${\mathrm{N}}_2{\mathrm{O}}_4\rightleftharpoons 2{ \mathrm{NO}}_2$$

(20)

Schmid mechanism

$${\mathrm{NO}}^{+}+{\mathrm{e}}^{-}\to \mathrm{NO}$$

(21)

$${\mathrm{HNO}}_2+{\mathrm{H}}^{+}\rightleftharpoons {\mathrm{NO}}^{+}+{ \mathrm{H}}_2\mathrm{O}$$

(22)

$${\mathrm{HNO}}_2+{\mathrm{HNO}}_3\to {\mathrm{N}}_2{\mathrm{O}}_4+{ \mathrm{H}}_2\mathrm{O}$$

(23)

$${\mathrm{N}}_2{\mathrm{O}}_4+2 \mathrm{NO} +2{ \mathrm{H}}_2\mathrm{O} \rightleftharpoons 4{ \mathrm{HNO}}_2$$

(24)

Another important pathway was proposed by Katsounaros et al. [44], who confirmed the formation of ${\mathrm{N}}_2$ and ${\mathrm{N}}_2\mathrm{O}$ from ${\mathrm{NH}}_2\mathrm{OH}$ in the presence of ${\mathrm{HNO}}_2$ and $\mathrm{HNO}$ (Equations (25) and (26)).

$${\mathrm{NH}}_2\mathrm{OH} + \mathrm{HNO}\to {\mathrm{N}}_2+2{ \mathrm{H}}_2\mathrm{O}$$

(25)

$${\mathrm{NH}}_2\mathrm{OH} +{ \mathrm{HNO}}_2\to {\mathrm{N}}_2\mathrm{O}+2{ \mathrm{H}}_2\mathrm{O}$$

(26)

In addition to the kind of nitrate reduction by the electron generated from the cathode as described above, nitrate can also be mediated by the atomic hydrogen (${\mathrm{H}}_{\mathrm{ads}}$) (Figure 3).

The process starts with the reduction in water ${\mathrm{H}}_2\mathrm{O}$ via the Volmer process, to form ${\mathrm{H}}_{\mathrm{ads}}$. Then, ${\mathrm{H}}_{\mathrm{ads}}$ which is considered a strong reducing agent (${\mathrm{E}}_{{\mathrm{H}}^{+}/\mathrm{H}}^0$ = −2.31 V vs. SHE) directly reduces nitrate to form the dominant end-product ${\mathrm{NH}}_4^{+}$. Notably, two ${\mathrm{N}}_{\mathrm{ads}}$ can combine to produce ${\mathrm{N}}_2$ but the formation of the N–N bond is kinetically less favorable than the N–H bond because the calculated migration barrier (Δ${\mathrm{E}}_{\mathrm{a}}$) of ${\mathrm{H}}_{\mathrm{ads}}$ is 0.10 eV, which is much lower than that of ${\mathrm{N}}_{\mathrm{ads}}$ (0.75 eV) [49]. The following electrochemical reactions described the mechanistic formation of ${\mathrm{N}}_2$ and ${\mathrm{NH}}_3$ (Equations (27)–(33)).

$${\mathrm{H}}_2\mathrm{O} +{ \mathrm{e}}^{-}\to {\mathrm{H}}_{\mathrm{ads}}+{ \mathrm{OH}}^{-}$$

(27)

$${\mathrm{NO}}_{3, \mathrm{ads} }^{–}+2{ \mathrm{H}}_{\mathrm{ads}}\to {\mathrm{NO}}_{2, \mathrm{ads} }^{–}+{ \mathrm{H}}_2\mathrm{O}$$

(28)

$${\mathrm{NO}}_{2, \mathrm{ads} }^{–}+{ \mathrm{H}}_{\mathrm{ads}}\to {\mathrm{NO}}_{\mathrm{ads}}+{ \mathrm{OH}}^{-}$$

(29)

$${\mathrm{NO}}_{\mathrm{ads}}+2{ \mathrm{H}}_{\mathrm{ads}}\to {\mathrm{N}}_{\mathrm{ads}}+{ \mathrm{H}}_2\mathrm{O}$$

(30)

$${\mathrm{N}}_{\mathrm{ads}}+{ \mathrm{H}}_{\mathrm{ads}}\to {\mathrm{NH}}_{\mathrm{ads}}$$

(31)

$${\mathrm{NH}}_{\mathrm{ads}}+{ \mathrm{H}}_{\mathrm{ads}}\to {\mathrm{NH}}_{2, \mathrm{ads}}$$

(32)

$${\mathrm{NH}}_{2, \mathrm{ads}}+{ \mathrm{H}}_{\mathrm{ads}}\to {\mathrm{NH}}_{3, \mathrm{ads}}$$

(33)

Based on the previously discussed reactions, it is evident that eNO₃R to produce ${\mathrm{N}}_2$ and ${\mathrm{NH}}_3$ involves intricate and multifaceted pathways. The main hurdle in this process lies in enhancing the kinetics of nitrite formation, known as RDS (rate-determining step). To overcome this challenge and optimize the overall performance of the process, it is imperative to gain a deeper understanding of the intricate mechanisms governing electrochemical nitrate reduction. This includes a comprehensive exploration of intermediates and reaction kinetics. Consequently, it becomes apparent that further investigation using advanced methodologies and theoretical studies is indispensable to validate the proposed reaction mechanism.

2.2. DFT Calculation as an Advanced Tool for eNO₃R

In recent years, significant advancements have been made in using Density Functional Theory (DFT) to simulate the mechanisms of nitrate reduction, providing new insights into this process. DFT calculations have revealed critical details about the adsorption energies of various nitrogen species (e.g., ${\mathrm{NO}}_3^{-}$, ${\mathrm{NO}}_2^{-}$, $\mathrm{NO}$, ${\mathrm{N}}_2$, etc.), which are necessary for assessing the catalytic activity of different materials. The adsorption of nitrates plays a crucial role in electrocatalytic processes as it represents the first step. Employing DFT calculation, it has been demonstrated that ${\mathrm{NO}}_3^{-}$ can be adsorbed on the active site via chemisorption (2 O and 1 O patterns) or physisorption (Figure 4) [50].

Furthermore, the presence of oxygen species at the catalytic sites is significant because it can facilitate this bond-breaking process, making it easier for the subsequent steps of nitrate reduction to occur. The subsequent steps of the mechanism following nitrate adsorption are generally exothermic, indicating the presence of multiple rate-determining steps within the process [51]. DFT, in conjunction with the use of descriptors, has proven instrumental in predicting novel materials by identifying specific RDS, as illustrated in

Table 2. Pathways and limiting steps for novel categories synthesizing electrode materials for eNO₃R using DFT calculations.

Electrode Materials	Free Energy Diagram	Rate Determining Steps	Refs.
Cu catalyst anchored on porous N-doped carbon. (Cu−NC)		*NO to *N	[52]
CuO(111) surface with two oxygen vacancy (OVs) (CuO−2OVs)		The protonation of *O-*N to*N-*OH	[53]
${\mathrm{MnCo}}_2{\mathrm{O}}_4$ (3 1 1) facet		*NO to *NOH	[54]
${\mathrm{MnCo}}_2{\mathrm{O}}_4$ (2 2 0) facet		*${\mathrm{N}}_2{\mathrm{O}}_2\mathrm{H}$ to *${\mathrm{N}}_2\mathrm{O}$	[54]
Au (111) (in yellow), Au (331) (in blue), Pd modified Au(331) (in red).		${\mathrm{NO}}_3^{-}$(ads) → ${\mathrm{NO}}_2^{-}$(ads) + O(ads) Au (111) = 0.72 eV Au (331) = 0. 39 eV Pd-Au (331) = 0.37 eV ${\mathrm{NO}}_2^{-}$(ads) → $\mathrm{NO}$ (ads) + O(ads) Au (111) = 1.74 eV Au (331) = 0.83 eV Pd-Au (331) = 0.23 eV	[55]
nanoporous ${\mathrm{Co}}_3{\mathrm{O}}_4/\mathrm{Co}$		*${\mathrm{NO}}_2$ to *$\mathrm{NO}$	[56]
Pd (in blue), $\mathrm{Pd}{({\mathrm{WO}}_3)}_3$ (in red).		*N to *NH Pd = 3.38 eV *${\mathrm{NH}}_2$ to *${\mathrm{NH}}_3$ $\mathrm{Pd}{({\mathrm{WO}}_3)}_3$ = 2.99 eV	[57]
Cu NWAs (in red),$\mathrm{Cu}/{\mathrm{Cu}}_2\mathrm{O}$ NWAs (in yellow)		The protonation of *O-*N to*N-*OH	[58]
${\mathrm{Ni}}_3{\mathrm{Co}}_6{\mathrm{S}}_8$ (NCS) (in blue) ${\mathrm{Ni}}_3{\mathrm{Co}}_6{\mathrm{S}}_8$ nanospheres with sulfur-rich vacancies (NCS-2) (in red)		*${\mathrm{HNO}}_3$ to *${\mathrm{NO}}_2$ NCS-2 = 0.24 eV NCS-2 = 0.19 eV	[59]
The incorporation of indium in sulfur-doped graphene (In-S-G)		${\mathrm{N}}_2$ formation pathway *NO to *${\mathrm{N}}_2{\mathrm{O}}_3$ (0.90 eV) ${\mathrm{NH}}_3$ production *${\mathrm{NO}}_3$ to *${\mathrm{NO}}_2\mathrm{OH}$ (0.63 eV)	[60]
Boron-doped copper nanowire electrocatalyst (B-Cu)		The hydrogenation of *NO to form H*NO	[61]
Cobalt manganese spinel nanoparticles (${\mathrm{CoMn}}_2{\mathrm{O}}_4)$		The desorption of *${\mathrm{NH}}_3$	[62]
Cobalt manganese spinel nanoparticles embedded in multichannel carbon fibers (${\mathrm{CoMn}}_2{\mathrm{O}}_4/\mathrm{NC}$)		*NO-to-*NHO	[62]
Iron coordinated with nitrogen on an ordered mesoporous carbon framework (Meso-Fe-N-C)		${\mathrm{N}}_2$ formation pathway ${\mathrm{N}}_2$ desorption ${\mathrm{NH}}_3$ production *NO-to-*NHO	[63]

.

This approach offers deeper insights into the selective catalytic reduction mechanisms of nitrates into gaseous nitrogen or ammonium. Similar advancements have been observed in hydrogen evolution reactions (HER), highlighting the competitive nature of these processes with eNO₃R. Furthermore, the integration of artificial intelligence (AI) and machine learning (ML) provides a promising avenue for predicting catalytic performance and significantly reducing the computational demands of DFT, thereby accelerating material discovery and optimization.

2.2.1. Insights into eNO₃R: Descriptors and Rate Determining Steps (RDS)

Descriptors are simplified parameters connected to fundamental surface properties that control both adsorption energies and activation energies for reactions. Moreover, a more suitable descriptor is the strength of surface-adsorbate bonds, which can be estimated from experimental data. This estimation facilitates the construction of a Sabatier-type ‘volcano’ plot, thus helping to identify optimal catalysts by correlating bond strength with catalytic activity [64]. For example, the binding energies of nitrate adsorbed (ΔG_*NO3) serve as the first descriptor in determining catalytic efficiency, as nitrate adsorption is recognized as the initial step involving a solution-mediated proton transfer without electron transfer, as previously described (This step has been previously identified as RDS) [65]. In addition to *${\mathrm{NO}}_3$, the formation of adsorbed *${\mathrm{NO}}_2$ and *$\mathrm{NO}$ represent additional key descriptors, with *${\mathrm{NO}}_2$ identified as the RDS and *$\mathrm{NO}$ as the potential-limiting step (PLS) in the process. Weak adsorption of ${\mathrm{NO}}_2$ and $\mathrm{NO}$ can result in their desorption into the solution, thereby negatively impacting the overall efficiency of the reaction. Figure 5A presents the binding energies of *${\mathrm{NO}}_3$, *${\mathrm{NO}}_2$, and *$\mathrm{NO}$ on different transition metals. The weak binding of *${\mathrm{NO}}_3$ and *${\mathrm{NO}}_2$ on Au results in poor catalytic performance for nitrate reduction. Moreover, most transition metals show a high desorption barrier for *$\mathrm{NO}$ except for Ag and Au [66]. To confirm these results, Karamad et al. [66] used the volcano framework as illustrated in Figure 5B, based on the Sabatier principle to analyze the catalytic activity of transition metals in the eNO₃R. They focus on the binding energies of nitrogen which correlate with the binding energies of *$\mathrm{NO}$ adsorbates (ΔG_*NO) and the binding energy of oxygen which correlates with *OH adsorbates (ΔG_*OH). They also identify the PLS as the reaction with the most negative limiting potential, revealing that for metals like Cu, Ni, Rh, Pt, and Pd, the key steps are *$\mathrm{NO}$ → *$\mathrm{NOH}$ or *$\mathrm{NO}$ → $\mathrm{H}$*$\mathrm{NO}$ which means that the protonation of *$\mathrm{NO}$ or *${\mathrm{NO}}_3$ is the slowest and most difficult step in the overall reaction pathway. Furthermore, Cu demonstrates superior catalytic activity despite a significant overpotential and is the most active and selective catalyst for eNO₃R to ammonia among the other transition metals (It will be useful in the discussion in Section 4 for copper-based electrocatalyst). They also proposed strategies for enhancing nitrate reduction catalysis including (a) alloying with oxygen-affinitive elements to stabilize intermediates like *$\mathrm{NOH}$ and $\mathrm{H}$*$\mathrm{NO}$, though care must be taken to avoid *$\mathrm{OH}$ poisoning; (b) tethering ligands that geometrically interact with *$\mathrm{NO}$/$\mathrm{H}$*$\mathrm{NO}$ to stabilize the latter without significantly affecting *$\mathrm{NO}$ binding; (c) employing promoters to modify adsorption sites and binding energies; and (d) leveraging hydrogen bonding with intermediates such as *$\mathrm{NHO}$/*$\mathrm{NOH}$ to achieve stronger stabilization and improved catalytic performance [66]. Another study by Liu et al. [67] predicted the activity of nitrate reduction electrocatalysts under different applied potentials (−0.2, 0, 0.2, and 0.4 V). They also constructed theoretical volcano plots using nitrogen binding energy (ΔE_N) and oxygen binding energy (ΔE_O) which serve as key descriptors for evaluating eNO₃R performance on metal surfaces, obtained via DFT-based mean-field microkinetic modeling. The volcano plots reveal how the nitrate reduction rate depends on the adsorption strengths of N and oxygen O species at four different applied potentials. RDS in nitrate reduction varies with applied potential and metal type. At low potentials (−0.2 to 0 V), the RDS for Pt and Pd is the dissociation of *${\mathrm{NO}}_3$. As the potential increases to 0.4 V, *$\mathrm{NO}$ desorption becomes rate-controlling for these metals. For Rh, *${\mathrm{N}}_2$ formation is rate-controlling above 0.2 V. In contrast, for Fe and Co, the RDS shifts from *${\mathrm{NO}}_3$ dissociation to N₂ formation as the potential surpasses −0.2 V. On Cu, the RDS transitions from *${\mathrm{NO}}_3$ dissociation to *${\mathrm{NO}}_2$ dissociation at 0.2 V due to increased *${\mathrm{NO}}_2$ coverage. For Ag and Au, *${\mathrm{NO}}_3$ dissociation remains the RDS across all potentials, attributable to their weak adsorption and low nitrate coverage [67]. As previously demonstrated, descriptors and volcano plots play a critical role in identifying RDS and determining optimal catalysts. These approaches have been extensively utilized to predict new catalysts and effectively distinguish between different materials based on their catalytic activity and selectivity, as exemplified by the study conducted by Shathishkumar et al. [68] who employed first-principles calculations to assess the electrocatalytic performance of dual-atom catalysts (DACs) based on ${\mathrm{TM}}_2/{\mathrm{N}}_6-\mathrm{G}$ (DACs supported by an N-doped graphene nanosheet where ${\mathrm{TM}}_2$ ranges from ${\mathrm{Sc}}_2$ to ${\mathrm{Zn}}_2$). Through high-throughput calculations, they identified ${\mathrm{Cr}}_2/{\mathrm{N}}_6-\mathrm{G}$, ${\mathrm{Mn}}_2/{\mathrm{N}}_6-\mathrm{G}$, and ${\mathrm{Cu}}_2/{\mathrm{N}}_6-\mathrm{G}$ as prominent candidates near the volcano plot as illustrated in Figure 5D, exhibiting impressive limiting potentials of −0.46, −0.45, and −0.36 V, respectively. Furthermore, significant energy barriers for byproduct formation (${\mathrm{NO}}_2$, $\mathrm{NO}$, and ${\mathrm{N}}_2\mathrm{O}$) which confirm their high selectivity [68].

2.2.2. Nitrate to Ammonium

For ammonia generation, research endeavors have shifted their focus towards investigating the subsequent step through DFT studies, specifically examining the hydrogenation of ${\mathrm{NO}}_{\mathrm{ads}}$. This pathway is widely acknowledged as the primary route for the synthesis of ammonium; however, it is complex to study due to the addition of hydrogen atoms, which can occur at both the nitrogen (N) and oxygen (O) atoms of $\mathrm{NO}$, two possible products can be formed: H$\mathrm{NO}$ (hydrogenated nitrogen) or $\mathrm{NO}$H (hydrogenated oxygen) [69]. Furthermore, recent studies have shown that the hydrogenation process occurs in four distinct steps. These steps correspond to different adsorption configurations of $\mathrm{NO}$, including O-end, O-side, N-end, and N-side pathways (Figure 6) [53]. These hydrogenation steps are also known for their significant energy barriers, which are influenced by several factors, with the electrocatalyst being the key determinant. Lim et al. [70] showed that the reduction in nitrate by different facets of Pd nanoparticles changes the selectivity and reduction rate. Energetics DFT calculations demonstrated that the reaction energy associated with the dissociation of *${\mathrm{NO}}_3$ to *${\mathrm{NO}}_2$ + *$\mathrm{O}$ was favorable on the Pd(111) facet than on the Pd(100) facet. In contrast, *${\mathrm{NO}}_2$ adsorbs more strongly on Pd(100) surfaces which favors strong ammonium formation compared to Pd(111) surfaces. Another study by Liu et al. [55] selected the Pd(111) plane as the Pd facet to build $\mathrm{Pd}{({\mathrm{WO}}_3)}_3$ catalyst and they have carried out DFT calculations to investigate the enhanced eNO₃R activity. The comparison between the calculated adsorption energy of ${\mathrm{NO}}_3^{-}$ on pure Pd and $\mathrm{Pd}{({\mathrm{WO}}_3)}_3$ showed that the energy barrier of the hydrogenation *N to *NH on pure Pd is the largest by 3.38 eV. However, the doping of the surface with ${({\mathrm{WO}}_3)}_3$ reduced this energy barrier. This study also reported that the reaction undergoes a *${\mathrm{NO}}_3$ → *${\mathrm{NO}}_2$ → *$\mathrm{NO}$ → *$\mathrm{N}$ → *$\mathrm{NH}$ → *${\mathrm{NH}}_2$ → *${\mathrm{NH}}_3$ pathway remaining the same but the RDS changed from the hydrogenation of *$\mathrm{N}$ to *$\mathrm{NH}$ on pure Pd to the hydrogenation of *${\mathrm{NH}}_2$ to *${\mathrm{NH}}_3$ on $\mathrm{Pd}{({\mathrm{WO}}_3)}_3$. Furthermore, it is important to note that RDS changes as a function of facet and material composition, as reported by Liu et al. [55].

2.2.3. Nitrate to Nitrogen

In the context of nitrogen formation, DFT computes the adsorption energy of nitric oxide on catalytic surfaces. Equally essential is the understanding of the intricate process of NO transfer from surface sites to the bulk solution, involving the assessment of energy profiles associated with desorption. Furthermore, the energetics of nitrogen–nitrogen (N–N) bond formation are of paramount significance [71]. The rational modification of catalyst structures can regulate the adsorption strengths between the active intermediates and the catalyst sites, suppressing competitive reactions and promoting the selectivity of the target product. Fan et al. [63] developed an electrocatalytic material called Meso-Fe-N-C, consisting of atomically dispersed iron coordinated with nitrogen on a mesoporous carbon framework, which greatly improves the selectivity of eNO₃R. They found that the generated *$\mathrm{NO}$, which is a key intermediate for the formation of ${\mathrm{N}}_2$ or ${\mathrm{NH}}_3$/${\mathrm{NH}}_4^{+}$, reacts with dissolved $\mathrm{NO}$ in solution to produce *${\mathrm{N}}_2\mathrm{O}$. Then, the latter can be easily reduced to *${\mathrm{N}}_2$ and desorbed from the active sites. They confirmed that the successive hydrogenation of *$\mathrm{NO}$ forms *$\mathrm{NO}$H, *${\mathrm{NH}}_2\mathrm{OH}$, and finally *${\mathrm{NH}}_3$. Moreover, the study concluded that the Fe–N active site is more thermodynamically favorable than Fe–Fe for ${\mathrm{N}}_2$ emergence. Furthermore, the confinement effect of mesoporous carbon is also a significant factor contributing to selective ${\mathrm{N}}_2$ generation, mainly through two aspects: geometry distortion of surface Fe–N–C active sites and modified the distribution of $\mathrm{NO}$ concentration in mesopores [63].

2.2.4. eNO₃R vs. HER

Generally, an ideal catalyst for eNO₃R should maintain high stability and catalytic activity while should effectively suppress the Hydrogen Evolution Reaction (HER). For this reason, Strong competition from HER presents a significant challenge in aqueous systems [72]. Notably, Hu et al. [51] found that HER competes strongly with nitrate reduction; at pH = 0, the formation of H₂ requires only 0.25 eV, lower than the 0.37 eV needed for nitrate reduction on Cu(111), making eNO₃R less favorable in acidic conditions. In neutral environments (pH = 7), eNO₃R outcompetes HER, while in alkaline conditions (pH = 14), HER predominates [51]. This indicates that pH significantly influences the competitive dynamics between HER and eNO₃R. However, it is not desirable to fully suppress HER, as the Adsorption strengths of the H atom or active hydrogen species are essential for the next step—nitrogen hydrogenation to produce ammonia, as discussed in the previous section of nitrate to ammonium. Typically, the binding energy of *H serves as a useful descriptor of hydrogen evolution activity across a variety of catalyst materials. To analyze selectivity trends for the NO3RR among transition metals, Karamad et al. [66] used an approach similar to that used for ${\mathrm{CO}}_2$ reduction. Accordingly, they plotted the difference U_L(eNO₃R) − U_L(${\mathrm{H}}_2$) against the limiting potential U_L(eNO₃R) for various transition metals as illustrated in Figure 7a. A more positive value of U_L(eNO₃R) − U_L(${\mathrm{H}}_2$) correlates with increased selectivity for eNO₃R over HER, suggesting that catalysts in the upper right region exhibit both high activity and selectivity for eNO₃R. Importantly, this analysis focuses on selectivity for the eNO₃R relative to HER rather than specifically towards ${\mathrm{NH}}_3$ or ${\mathrm{N}}_2$. The results suggest that Cu emerges as the most active and selective catalyst for the eNO₃R, aligning with previous studies identifying Cu as a leading candidate. Additionally, while Ir shows higher catalytic activity than Cu, its lower selectivity towards eNO₃R products leads to ${\mathrm{H}}_2$ being the dominant product under eNO₃R conditions. Furthermore, it is critical to develop strategies to effectively manage this competition. Therefore, active sites should preferentially bind NO₃ over H atoms, with ΔG_NO3 > ΔG_H, and the formation of HER byproducts can be minimized with U_L(${\mathrm{H}}_2$) > U_L(${\mathrm{NH}}_3$) [66]. In this context, Sathishkumar et al. [68] considered ${\mathrm{TM}}_2/{\mathrm{N}}_6-\mathrm{G}$ (${\mathrm{TM}}_2$ = ${\mathrm{Sc}}_2$, ${\mathrm{Ti}}_2$, ${\mathrm{V}}_2$, ${\mathrm{Cr}}_2$, ${\mathrm{Mn}}_2$, ${\mathrm{Fe}}_2$, ${\mathrm{Co}}_2$, ${\mathrm{Cu}}_2$, and ${\mathrm{Zn}}_2$) as potential catalysts. The adsorption capabilities of ${\mathrm{NO}}_3$ and the suppression of HER (illustrated in Figure 7b) was assessed. Among the nine candidates, ${\mathrm{Cr}}_2/{\mathrm{N}}_6-\mathrm{G}$, ${\mathrm{Mn}}_2/{\mathrm{N}}_6-\mathrm{G}$, and ${\mathrm{Cu}}_2/{\mathrm{N}}_6-\mathrm{G}$ met the selectivity criteria: ΔG_NO3 > ΔG_*H and U_L(${\mathrm{H}}_2$) > U_L(${\mathrm{NH}}_3$). The free energy diagrams for HER indicated uphill ΔG_PDS values of 0.59, 0.24, and 0.37 eV for ${\mathrm{Cr}}_2/{\mathrm{N}}_6-\mathrm{G}$, ${\mathrm{Mn}}_2/{\mathrm{N}}_6-\mathrm{G}$, and ${\mathrm{Cu}}_2/{\mathrm{N}}_6-\mathrm{G}$, respectively, at their corresponding limiting potentials of −0.46 V, −0.45 V, and −0.36 V. Thus, these three systems were identified as promising candidates for nitrate reduction to ammonia [68].

2.2.5. Analyzing Experimental Data vs. DFT for Nitrate Reduction

The comparison between experimental data and DFT-based simulations reveals notable discrepancies in the detailed energetics of intermediate steps within eNO₃R. Regarding that, Long et al. [73] reported that product selectivity and Faradaic efficiencies (FE) for ${\mathrm{N}}_2\mathrm{O}$, ${\mathrm{NH}}_3$, and ${\mathrm{H}}_2$ obtained from a nanostructured Ag electrode in phosphate-buffered saline (PBS) show distinct potential windows, where ${\mathrm{NH}}_3$ production becomes dominant at lower potentials, and N₂O prevails at higher potentials. This experimental observation aligns with predictions from DFT-based microkinetic simulations, as shown in Figure 8a,b, which highlight potential-dependent shifts in selectivity. However, discrepancies in the energetics of intermediate steps arise; while DFT predicts the RDS for ${\mathrm{NH}}_3$ production as the electrochemical protonation of H*NOH, experimental results indicate deviations in the exact potential for product transitions. Similarly, the competition between ${\mathrm{NH}}_2\mathrm{OH}$ and ${\mathrm{NH}}_3$ production is captured by DFT, though slight variations in ${\mathrm{NH}}_2\mathrm{OH}$ suppression onset are noted experimentally [73]. Despite these differences, microkinetic modeling plays a crucial role in bridging the gap between theory and experiment. As shown by Carvalho et al. [74] who compared experimental observations with model predictions to elucidate nitrate reduction reaction (NRA) dynamics on transition metal surfaces (Figure 8c,d). The experimental rate-order profiles exhibited a potential-dependent peak, characteristic of a competitive Langmuir–Hinshelwood mechanism, indicating interdependence between H* and nitrate adsorption. By calibrating a microkinetic model to incorporate the rate-limiting steps of nitrate reduction and hydrogen evolution kinetics, they achieved agreement between experimental and modeled nitrate rate orders by adjusting thermodynamic parameters (ΔG_H*–ΔG_NO3⁻). The model effectively captured observed trends, including shifts in peak rate order and mass transfer limitations, reinforcing its validity and providing key insights into the competitive adsorption and reaction kinetics across various transition metal catalysts [74]. Furthermore, the integration of machine learning and Artificial Intelligence techniques holds significant potential to further facilitate these modeling efforts by optimizing parameter selection and improving predictive accuracy. This will be explored in the next section.

2.3. Machine Learning and Artificial Intelligence for Efficient eNO₃R

For eNO₃R, computational methods like DFT, molecular dynamics, and microkinetics have been invaluable in predicting catalyst behaviors and screening materials as discussed above. However, traditional atomistic simulations, while accurate, are computationally intensive, limiting their ability to explore large catalyst design spaces. Machine learning (ML) offers a solution by rapidly predicting catalyst performance using scaling relationships like Brønsted–Evans–Polanyi (BEP) and adsorption-energy trends. By leveraging data-driven approaches, ML can identify optimal catalytic properties for eNO₃R [75]. As demonstrated by Chen et al. [76] ML models like Gradient Boosting Regression (GBR) were trained using key intrinsic properties of transition metals—such as d-electron number, electronegativity, and atomic radius—to predict adsorption energies of nitrogen intermediates. This approach significantly reduced computational costs, offering reliable predictions for both side-on and end-on adsorption modes, with the potential to accelerate the discovery of highly efficient eNO₃R catalysts. The model’s interpretability further enhances the understanding of catalyst behavior, highlighting the role of ML in advancing catalytic design for nitrate reduction [76]. In addition, a study by Liu et al. [77] discovered that there was no linear correlation between the adsorption-free energies of reaction intermediates, making traditional free energy descriptors ineffective. For this reason, they employed Random Forest (RF) algorithm to predict the limiting potentials (U_L) for electrocatalytic NO reduction (NORR) to ${\mathrm{NH}}_3$, utilizing 22 atomic, electronic, and custom features. Through Pearson’s correlation, 11 key features were selected for further analysis, including atomic structure and bond lengths. The RF model demonstrated excellent performance with an R² value of 0.96 and low RMSE, highlighting important features such as bond length, atomic number, and d-band center in influencing U_L. Furthermore, they applied the SISSO approach (Sure Independence Screening and Sparsifying Operator) to derive a simplified mathematical expression that links U_L to the key features, which aligned well with DFT predictions. As a result, Cu and Pt/Cu single-atom catalysts (SAACs) were identified as promising candidates for NORR to ${\mathrm{NH}}_3$. [77]. Another study by Yang et al. [78] applied multiple ML algorithms—SVR, GPR, LASSO, RFR, and XGBR—to further investigate the structure-activity relationship in NORR. A key component of this analysis involved the use of a normalized dataset, with a 10-fold cross-validation approach used to identify optimal hyperparameters for the ML models. The Random Forest Regression (RFR) and XGBR models outperformed others, demonstrating excellent generalization ability with R² values of 0.95 and 0.98, and RMSE values of 0.30 eV and 0.19 eV, respectively. These models provided reliable predictions of Gibbs free energy changes for the reaction intermediates [78].

Other models have been developed to improve the prediction of adsorption energies and active sites, such as the TinNet (Theory-Infused Neural Network) model and the Bayesian Statistical Learning model (Bayeschem). TinNet integrates traditional theoretical principles with neural networks (AI-based machine learning) which integrates domain-specific physical theories, including the d-band theory of chemisorption, into machine learning frameworks, particularly graph neural networks (GNNs). By embedding these physical principles, TinNet effectively captures complex interactions between atoms and surfaces, resulting in more accurate predictions of adsorption energies and chemical reactivity. On the other hand, the Bayeschem model used a statistical approach to optimize relationships between variables, like material composition and electronic properties, through global optimization techniques such as Markov Chain Monte Carlo (MCMC) [75]. For instance, Gao et al. [79] employed the Bayeschem model to analyze the adsorption properties of *$\mathrm{N}$ and *${\mathrm{NO}}_3$ on selected pristine metal surfaces. Their study revealed a mechanism to bypass linear adsorption-energy scaling limitations by exploiting site-specific Pauli repulsion interactions. This work highlighted how an enhanced understanding of the interactions between adsorbates and catalyst surfaces can lead to improved electrocatalytic performance. Ultimately, the Bayeschem model plays a crucial role in bridging theoretical insights with practical applications, facilitating the design of more efficient electrocatalysts for nitrate reduction [79].

By employing ML, researchers can efficiently evaluate and optimize new catalysts for electrochemical nitrate reduction, thereby minimizing the reliance on costly simulations and facilitating a more effective discovery of high-performing catalytic solutions [80].

3. Recent Developments in Electrocatalyst Fabrication and Characterization Techniques for Effective eNO₃R Process

Currently, numerous techniques including electrodeposition, plasma treatment, thermal treatment, and so on, have been developed for the synthesis of diverse electrocatalysts for nitrate reduction, which plays a critical role in determining the purity and homogeneity of the electrode surface, which are critical parameters influencing the eNO₃R efficiency and the catalyst’s selectivity. Furthermore, there are several well-known methods for material characterization, which can be used to characterize electrocatalysts both before and after reactions, including SEM, TEM, XRD, and XPS. Notably, in situ characterization techniques are particularly important for observing changes during electrolysis. Therefore, in this section, we will discuss the different types of electrode preparation methods, followed by an overview of advanced in situ characterization techniques, highlighting their advantages and limitations.

3.1. Advances in Electrocatalyst Preparation Techniques for Efficient Nitrate Reduction

3.1.1. Electrodeposition

Electrodeposition involves depositing material onto an electrode by applying an electrical current or potential in an electrochemical cell. The performance of electrodeposited electrocatalysts is highly dependent on the specific electrodeposition parameters used. To optimize these electrocatalysts, a detailed investigation is needed to understand how different parameters affect the deposition process [81]. Key factors, such as the choice of electrolyte, electrodeposition time, mode of electrodeposition, and the presence of additives, play a crucial role in shaping the performance of the deposited electrocatalyst. Regarding that, Zurita et al. [82] investigated how the electrolyte in the electrodeposition process affects the morphology of copper nanoparticles, specifically on Vitreous Carbon (VC) substrate, observing that ${\mathrm{Na}}_2{\mathrm{SO}}_4$ solution led to flower-shaped dendritic structures; however, ${\mathrm{H}}_2{\mathrm{SO}}_4$ produced randomly distributed hemispherical particles. The differences in morphology on VC were linked to substrate defects interacting with the electrolytic media, affecting the nucleation and growth of the crystals [82]. Another factor which is the electrodeposition time, plays a crucial role in the performance and electrocatalysts stability for nitrate reduction. Specifically, longer deposition times can lead to issues with electrode stability, as seen in the case of Cu-BDD electrodes. For instance, a copper oxide layer formed after 600 s of electrodeposition was prone to detachment during electrochemical nitrate reduction. This instability can hinder the electrocatalyst’s performance and longevity. In contrast, a deposition time of 300 s was found to be optimal. At this duration, copper particles formed on the Cu-BDD surface were unevenly distributed with diameters ranging from 72.7 to 139 nm, and an average size of 106 nm. The shorter deposition time resulted in better particle adhesion and electrode stability, making it more suitable for eNO₃R [83]. Another study by Zhang et al. [84] demonstrated that longer electrodeposition times result in higher Cu loading on the electrode, enhancing ${\mathrm{NO}}_3^{-}$ removal due to increased catalytic activity. However, after 15 min, Cu is evenly distributed on the Ni foam surface, and additional electrodeposition time does not significantly increase active sites. Additionally, extended electrodeposition times (25 and 35 min) can lead to Cu agglomeration, which reduces the degradation rate of ${\mathrm{NO}}_3^{-}$ [84]. The advantage of the electrochemical deposition method is its ability to synthesize electrocatalysts with specific crystallographic facets tailored to needs. For example, Chen et al. [85] synthesized Tetrahexahedral Cu nanocrystals (THH Cu NCs) with {5 2 0} facets using a programmed square-wave potential (SWP) method, which allows precise control over the exposed crystallographic facets by adjusting potential limits. The synthesized THH Cu{5 2 0} nanocrystals demonstrated exceptional catalytic activity and selectivity in the electrochemical reduction in nitrate, achieving a Faradaic efficiency of 98.3% at −0.90 V (vs. RHE) in neutral solution [85]. However, the drawback of this electrocatalyst preparation method is the instability of the electrodes. According to Chen et al. [86] who developed Pd/Cu electrocatalysts with electrodeposition durations between 60 and 360 s, Pd/Cu 180 s demonstrated superior nitrate removal efficiency and nitrogen selectivity. However, ICP analysis identified a Pd concentration of 0.063 ppm after prolonged reaction times, highlighting the necessity for future improvements in electrode stability, possibly by adding other treatments [86].

3.1.2. Plasma Treatment

Plasma treatment uses an ionized gas to modify the surface properties of the materials and specifically to control oxygen vacancies in electrocatalyst materials [87]. different plasma treatments with ${\mathrm{N}}_2$, Ar, and ${\mathrm{O}}_2$ are employed to achieve high and low concentrations of oxygen vacancies in the samples. Jang et al. [88] used a GVTech GVPT–100 plasma system, with a 100 W power supply and at a pressure of 180 Pa. ${\mathrm{MnCuO}}_{\mathrm{x}}$ samples were exposed to Ar and ${\mathrm{O}}_2$ plasma for 240 s, generating different concentrations of oxygen vacancies, with Ar creating a higher concentration and ${\mathrm{O}}_2$ a lower concentration. Their study confirms that plasma treatment modifies the oxygen vacancy concentration of electrocatalysts without altering morphology or bulk structure. Furthermore, This concentration of oxygen vacancies has a significant effect in improving the adsorption energy of nitrate and intermediates which significantly improves the ${\mathrm{NH}}_3$ yield rate and FE as illustrated in Figure 9 [88].

3.1.3. Thermal Treatment (Calcination)

The calcination treatment requires thermal treatment at controlled temperatures for a defined time [89]. Zhang et al. [90] prepared a Copper-modified titanate cathode (Cal-Ti/Cu) at 350 °C for 30 min for nitrate reduction in boiling water. They demonstrated that Nitrate reduction efficiency with Cal-Ti/Cu (65.37%) is much higher than Ti (5.3%) and Ti/Cu (5.21%) at 80 °C. Furthermore, the calcination of electrocatalysts increases surface area, active sites, and corrosion resistance, preventing surface deactivation [90]. Another study by Liu et al. [91] prepared Ni foam/Carbon nanotubes (CNTs)/Cu electrodes at different temperatures (350 °C, 400 °C, and 450 °C). The maximum nitrate removal was achieved at 450 °C due to better binding of CNTs to Ni foam and improved CNT film properties with this specific temperature of calcination [91].

3.1.4. Thermal Treatment (Annealing)

Thermal treatment (annealing) involves heating a material to a precise temperature for a defined period and then cooling it under controlled conditions. Annealing is generally used in the middle of catalyst preparation and specifically for catalyst decomposition such as the conversion of Cu(OH)₂ to CuO [92]. Electrocatalyst materials subjected to annealing showed significant improvements in nitrate reduction performance compared to those without annealing. Kuang et al. [83] showed that annealing increased BDD’s surface energy, allowing better copper attachment due to numerous dangling bonds, additionally the annealing process consolidated copper oxides on the BDD surface, which significantly enhanced electrode stability. The surface morphology of the BDD electrode remained unchanged before and after copper oxide removal, indicating its suitability for long-term use. Furthermore, the BDD electrode could be reused multiple times without significant changes in copper oxide morphology [83].

3.1.5. Hydrothermal Treatment

The hydrothermal treatment method is generally used in a high-temperature and high-pressure aqueous environment with an autoclave to grow or synthesize the specific size and shape of a monodispersed catalyst [93]. This treatment is generally followed by other specific treatments, as illustrated in the work of Chen et al. [94] who used immersed carbon cloth in a solution containing $\mathrm{Bi}{({\mathrm{NO}}_3)}_3·5{\mathrm{H}}_2\mathrm{O}$ and then subjected to hydrothermal treatment in an autoclave with an initial impregnation step in a solution containing metal precursors ($\mathrm{Cu}{({\mathrm{NO}}_3)}_3·3{\mathrm{H}}_2\mathrm{O}$) to incorporate the metal into the material followed by annealing in the 350 °C muffle furnace for 10 min, to synthesize ${\mathrm{Cu}}_2\mathrm{O} \mathrm{NPs}/{\mathrm{BiO}}_{2-\mathrm{x}}$nanosheets. Hydrothermal treatment has a great influence on improving the oxygen vacancy-rich BiO₂-x which enhances eNO₃R efficiency [94].

3.1.6. Impregnation Method

The preparation of the electrocatalyst materials using the impregnation method requires immersing the noble metal in a solution, followed by drying and calcination to complete the process [95]. The impregnation method is commonly used to optimize the amount of noble metal to be added in order to minimize the loss of noble metals, as demonstrated in the study by Zhang et al. [96] who synthesized $\mathrm{Pd}–\mathrm{Cu}/{\mathsf{\gamma }\mathrm{Al}}_2{\mathrm{O}}_3$ catalysts with different metallic palladium loading rates. As reported in their study, a palladium loading rate of 5% demonstrated the best performance among the various tested rates and was, therefore, selected for subsequent experiments [96].

3.2. Advanced in Situ Characterization Techniques

Electrode materials gained particular attention for their ability to ensure the eNO₃R mechanism. Hence, to establish a link between the electrode surface structure, (including its chemical composition, crystal structure, coordination number, and interatomic distance) and the eNO₃R performance, in situ characterization techniques have been employed during the electrochemical reaction. These techniques mainly include X-ray absorption spectroscopy (XAS), X-ray photoelectron spectroscopy (XPS), in situ electrochemical scanning tunneling microscopy (EC-STM), grazing-incidence X-ray diffraction (GIXRD), and shell isolated nanoparticle enhanced Raman spectroscopy (SHINERS) [57,97,98,99,100,101]. These techniques allow the real-time control of the reaction and the characterization of the catalysts under reaction conditions. Thanks to the use of these techniques, a qualitative understanding of the eNO₃R reaction mechanism is possible. In addition, they can provide information on dynamic changes in the catalyst’s stereo geometry and electronic structure. Thus, the development of in situ techniques is of utmost importance to identify the active species involved in the eNO₃R mechanism and provide a qualitative understanding of the reaction mechanism. The various in situ techniques are summarized in the following

Table 3. In situ Characterization Techniques for eNO₃R.

Characterization Techniques	Detected Products and Intermediates	Principles	Advantages	Disadvantages	Refs.
Shell isolated nanoparticle enhanced Raman spectroscopy (SHINERS)	All the intermediates adsorbed on different faces of catalyst material.	The principle is based on measuring the inelastic scattering of light caused by molecular vibrations involving changes in the polarization of the molecule. It consisted of a gold nanoparticle with an insulating thin SiO2 shell (Au/SiO2) to eliminate the interference of deposited Au or other substrates (Ag, Au), which have a surface plasmon resonance effect, and to inhibit the signal from the analytes adsorbed on the Au core.	High sensitivity, rapid, less destructive, and noninvasive technique.	It cannot be used for the detection of atomic cations and anions directly, complicated process, and has a weak pH tolerance.	[102,103]
Fourier-transformed infrared spectroscopy (FT-IR)	All intermediates formed at the surface of the electrode.	The incident infrared ray is guided by the prism and optical stage toward the chemical bonds of intermediates formed at the surface of the electrode, which was then diffused and reflected in the detector. The FT-IR measurements were performed on a spectrometer equipped with a liquid-nitrogen-cooled detector.	Rapid determination of adsorbed intermediates, applied in wide concentration ranges, nondestructive, and repeatable.	FTIR instruments have only a single beam.	[104,105,106]
X-Ray Absorption Spectroscopy	Active sites at the surface of the electrode.	The method involves selecting the desired energy of X-ray radiation using a monochromator and measuring the resulting XAS spectrum, which is composed of two parts: the X-ray absorption near-edge structure (XANES) and the extended X-ray absorption fine structure (EXAFS). The XANES spectrum provides information on the electronic structure of the element being studied, while the EXAFS spectra give information on the local geometry of the atom. This technique is well suited for identifying active sites in multi-element systems and can measure elements at low concentrations due to their high flux properties.	No crystal and ordered structure is needed for the sample additionally the method works for single-atom materials too.	XAS data represents an average of the entire sample and Soft XAS requires an ultra-high or high vacuum environment. XAS may not provide detailed information about specific regions of the sample.	[107,108,109]
Electrochemical Scanning Tunneling Microscopy	surface morphology	In situ, electrochemical scanning tunneling microscopy (EC-STM) is a form of scanning probe microscopy that uses a metal needle tip and an electrochemical control circuit to acquire high-resolution images of a sample surface. It operates on the quantum tunneling effect and can work in various environments, including ultra-high vacuum and liquid environments.	Applicable to a wide range of environments, Wide working temperature range, and high resolution.	Damage risk in constant height mode on undulating surfaces, cannot detect bulk structure information of the sample, cannot directly observe insulator materials, and the resolution of the image to the real surface is limited if the sample surface is covered with a conductive layer.	[110,111]

, which outlines the detected products and intermediates, principles, advantages, and disadvantages of each method.

4. Advanced Strategies for eNO₃R Toward Ammonia and Nitrogen

eNO₃R process has been extensively studied using a wide range of materials, including platinoids which have been used several times for nitrate reduction such as Pt [112,113,114,115], Ru [116], Pb [117], Rh [118,119], Ir [120]. Thus there are other materials that have been studied such as Sn [121,122], Zn [117], Ni [115,117]. Additionally, many researchers have studied alloys containing at least two components and having to contain at least one metal, such as platinoid alloys including Pt-Ir [120], Pt-Sn [123,124,125], Pd-Sn [125,126,127], or Pt-Rh [128]. Furthermore, non-metallic materials, including graphene, graphite, [129,130], carbon fiber [131], Boron doped diamond (BDD) [132,133], and so on, have emerged as promising catalysts for eNO₃R, offering advantages such as low-cost, durability, and tunable electronic structures. In the search for efficient eNO₃R catalysts, researchers’ efforts diverge to synthesize materials effective in reducing nitrates to nitrogen and ammonium. Figure 10 and

Table 4. Ammonia production with different Cu-based materials and nitrogen selectivity with various Fe-based catalysts.

Cathode	Ammonia Production		Nitrogen Selectivity		Refs.
	Conditions	Performance	Conditions	Performance
Copper-Based Electrocatalysts
Cu Single Atom Gel (SAG)	0.1 M phosphate buffer solution (PBS), pH 7, 20 mM KNO3, and Overpotential −0.9 V vs. Reference Hydrogen Electrode (RHE)	${\mathrm{NH}}_3$ yield rate: 0.53 mg cm−2 h−1; FE: 95.3%	-	-	[101]
Cu-N-C SAC Copper into porous nitrogen-doped carbon	0.1 M KOH, 0.1 M KNO3, and Overpotential − 1.00 V vs. RHE	${\mathrm{NH}}_3$ yield rate: 4.5 mg cm−2 h−¹; FE: 84.7%	-	-	[135]
Cu nanodisks (111)	0.1 M KOH, 10 mM KNO3, 10 mM KNO2 and Overpotential −0.5 V vs. RHE	${\mathrm{NH}}_3$ yield rate: 2.18 mg mgcat−1 h−¹; FE: 81.1%			[136]
Cu nanosheets	0.1 M KOH 170 ppm NO3—N and Overpotential −0.15 V vs. RHE	${\mathrm{NH}}_3$ production rate: 0.39 mg mgcat−1 h−¹; FE: 99.7%	-	-	[137]
Au/Cu	0.5 M NaNO3, 1 M NaOH, and Overpotential −0.7 V vs. RHE	${\mathrm{NH}}_3$ yield rate: 73.4 mg cm−2 h−1; FE: 98.02%	-	-	[138]
Cu50Co50	1 M KOH, 100 mM KNO3, and Overpotential −0.2 V vs. RHE	${\mathrm{NH}}_3$ yield rate: 81.7 mg cm−2 h−¹; FE: 100 ± 1%	-	-	[139]
Cu/Cu2O	0.5 M Na2SO4 14.3 mM ${\mathrm{NO}}_3^{-}$ and Overpotential −0.85 V vs. RHE	${\mathrm{NH}}_3$ production rate: 4.088 mg cm−2 h−1; FE: 95.8%	-	-	[58]
Cu2O-Cu/Ti	0.1 M KNO3, 1 M KOH, and Overpotential 0.5 V vs. RHE	${\mathrm{NH}}_3$ yield rate: 4.76 mg cm−2 h−1; FE: 92%	-	-	[140]
Cu/CuOx/Co/CoO	0.1 M ${\mathrm{NO}}_3^{-}$, pH 13, 0.1 M KNO3, and Overpotential −0.175 V RHE	${\mathrm{NH}}_3$ yield rate: 19.94 mg cm−2 h−1; FE: 93.3%	-	-	[141]
Iron-based electrocatalysts
Fe	-	-	3.5 mM Na2SO4 + 7 mM NaNO3 electrolysis duration of 3 h.	87% nitrate reduction efficiency with 100% nitrogen selectivity.	[142]
nZVI@OMC mesoporous carbon supported nanoscale zerovalent iron	-	-	50 mg/L ${\mathrm{NO}}_3^{-}$– N;0.02 M NaCl and 24 h of electrolysis.	65% nitrate conversion with 74% nitrogen selectivity.	[143]
nZVI@D201 polymeric beads supported by nZVI	-	-	50 mg/L ${\mathrm{NO}}_3^{-}$–N; 1.0 M NaCl and 60 h of electrolysis.	80% nitrate conversion with 95% nitrogen selectivity.	[144]
FeNC/MC Iron over N-doped mesoporous Carbon	-	-	100 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.1 M Na2SO4 and 24 h of electrolysis.	87% nitrate conversion with 81% nitrogen selectivity.	[145]
Fe#OMC Fe(0)-dispersed ordered mesoporous carbon	-	-	50 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.1 M Na2SO4; 0.02 M NaCl and 24 h of electrolysis.	86.9% nitrate conversion with 100% nitrogen selectivity.	[146]
Fe@PMO Iron Nanoparticles Confined in Periodic Mesoporous Organosilicon	-	-	100 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.02 M Na2SO4; 0.02 M NaCl and 24 h of electrolysis.	90% nitrate conversion with 99% nitrogen selectivity.	[147]
FeN-NC N-doped Fe nanoparticles	-	-	100 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.1 M Na2SO4; 0.02 M NaCl and 24 h of electrolysis.	91% nitrogen selectivity	[148]
Fe@N-C N-doped graphitic carbon-encapsulated iron nanoparticles	-	-	50 mg/L ${\mathrm{NO}}_3^{-}$–N; 1.0 g/L NaCl and 24 h of electrolysis.	83% nitrate conversion with 100% nitrogen selectivity	[149]
FeNi/g-mesoC/NF binderless FeNi/graphitized mesoporous carbon on Ni Foam	-	-	50 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.05 M Na2SO4; 0.02 M NaCl and 24 h of electrolysis.	75% nitrate conversion with 100% nitrogen selectivity	[150]
Fe3Ni-N-C Fe/Ni bimetallic nitrogen-doped porous carbon	-	-	100 ppm ${\mathrm{NO}}_3^{-}$-N; 0.1 M Na2SO4 and 0.5 h of electrolysis.	97.9% nitrate conversion with 99.3% nitrogen selectivity.	[151]
Fe0/Ni2P/CC	-	-	15 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.3 M Na2SO4 and 4 h of electrolysis.	89.81% nitrate conversion with 95.55% nitrogen selectivity.	[152]
Ni-Fe0@Fe3O4 Fe0@Fe3O4 nanoparticles immobilized on nickel foam	-	-	0.5 g/L Cl-, 50 mg/L ${\mathrm{NO}}_3^{-}$, Current density of 5 mA/cm2, electrolysis duration of 4 h, and pH of 6.2.	90.19% nitrate conversion with 88.09% nitrogen selectivity.	[153]
CL-Fe@C Iron Nanoparticles in Carbon Microspheres	-	-	100 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.02 M NaCl and 48 h of electrolysis.	54% nitrate conversion with 98% nitrogen selectivity.	[154]
CC/Fe@C carbon-coated iron nanoparticles on a functionalized carbon cloth	-	-	50 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.01 M Na2SO4; 0.01 M NaCl and 24 h of electrolysis.	92% nitrate conversion with 82% nitrogen selectivity.	[155]
B-Fe NCs boron-iron nanochains	-	-	100 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.02 M Na2SO4; 0.02 M NaCl and 24 h of electrolysis.	80% nitrate conversion with 99% nitrogen selectivity.	[156]
Fe@C-1 iron nanoparticles uniformly embedded in the carbon microsphere	-	-	100 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.02 M NaCl and 48 h of electrolysis.	75.9% nitrate conversion with 98% nitrogen selectivity	[157]
Fe/Fe3C-NCNF-2 Fe/Fe3C nanoparticle-decorated N-doped carbon nanofibers			100 mg/L ${\mathrm{NO}}_3^{-}$–N; 0.02 M NaCl and 24 h of electrolysis.	95% nitrate conversion with 100% nitrogen selectivity.	[158]

showed that copper-based catalysts have demonstrated interesting results in yield rates and faradaic efficiency (FE) compared to the value required for practical application (∼57 mg cm⁻² h⁻¹, faradaic efficiency FE > 90%) [134]. Various strategies have been explored to enhance Cu-based catalysts including facet manipulation, defect engineering, geometric regulation, fabricating copper composites with other metals or non-metals, bimetallic catalysts, encapsulation of active species, nanostructuring, and so on. Therefore, in the following section, the strategies used to convert nitrate to non-toxic gaseous nitrogen, in particular with copper and iron nanomaterials, will be outlined. As a matter of fact, these latter have yielded interesting results, as shown in Table 4, with the emphasis on their advantages. The engineering of zero-valent iron to advance the process is particularly addressed. Additionally, the strategy to convert ammonia generated through indirect oxidation by chloride-electrochemical advanced oxidation is discussed.

4.1. Copper-Based Electrocatalyst for the eNO₃R: A Case Study for Ammonia Production

4.1.1. Copper-Based Single Atom Catalyst (Cu-Based SACs)

Cu-based SACs have demonstrated exceptional performance in the eNO₃R owing to the intrinsic electronic configuration of Cu, which facilitates strong interaction with nitrogen oxyanions such as ${\mathrm{NO}}_3^{-}$. For instance, Cu incorporated into 3,4,9,10-perylenetetracarboxylic dianhydride (Cu-PTCDA) exhibits high selectivity (FE = 85.9%) and a substantial production rate (436 $\mathsf{\mu }\mathrm{g}.{\mathrm{h}}^{-1}.{\mathrm{cm}}^{-2}$) for ${\mathrm{NH}}_3$, surpassing other transition metals (Ag, Bi, Ir, Pt, Co, Fe, Ni). The Cu atom in Cu-PTCDA forms strong bonds with ${\mathrm{NO}}_3^{-}$, promoted by the efficient orbital overlap between ${\mathrm{NO}}_3^{-}$’s 2p orbitals and Cu’s d orbitals [159]. Cu incorporated into nitrogen-doped carbon (Cu-N-C) further enhances eNO₃R activity, particularly Cu-N-C-800, which shows superior NH₄⁺–N selectivity (80%) compared to bulk Cu. The Cu-N-C-800 catalyst features Cu–N₂ and Cu–N₄ moieties that exhibit high adsorption energies for ${\mathrm{NO}}_3^{-}$ and ${\mathrm{NO}}_2^{-}$, crucial for efficient eight-electron transfer in the eNO₃R. This high adsorption capability enables Cu-N-C-800 to reduce ${\mathrm{NO}}_3^{-}$ to ${\mathrm{NH}}_2^{-}$ without releasing ${\mathrm{NO}}_2^{-}$ into the electrolyte, highlighting its role in accelerating ${\mathrm{NH}}_3$ production [160]. Contrasting Cu SACs with other transition metals, SACs reveal Cu’s unique advantages in the eNO₃R. While Fe SACs also show promising performance, particularly in terms of ${\mathrm{NH}}_3$ FE and yield rate, Cu SACs exhibit superior thermodynamic properties and kinetics due to stronger ${\mathrm{NO}}_3^{-}$ coordination with Cu–N₄ moieties [161]. Cu-based SACs represent a promising avenue for advancing electrocatalysts in the eNO₃R, leveraging their unique electronic properties and strong interactions with nitrogen species to achieve high selectivity and efficiency in ${\mathrm{NH}}_3$ production.

4.1.2. Facet Engineering

Copper-based catalysts exhibit notable catalytic activity, especially when possessing a special electronic structure. Among these, Cu has demonstrated the highest activity, primarily yielding ammonia as the main product. The performance of Cu electrodes in nitrate electroreduction is closely linked to their structural characteristics. Additionally, the specific atomic structure of Cu catalysts, such as Cu(100) and Cu(111) surfaces, significantly influences their electrochemical behavior. In alkaline media, Cu(100) is more active in forming hydroxylamine compared to Cu(111), and the selectivity for nitrate reduction products can be controlled by adjusting the ratio of Cu(200) to Cu(111), which can be further tuned using surfactants to optimize nitrogen selectivity [162,163,164]. Studies on Cu₂O electrocatalysts have shown that manipulating exposed facets, particularly the (100) facet, enhances ammonia yield compared to other facets like (111), demonstrating the importance of surface structure in catalytic performance [165]. Bae and Gewirth [166] compared eNO₃R activity on Cu(111) and Cu(100) surfaces in acidic electrolytes. As a result, Cu(111) demonstrated superior activity thanks to the easier formation of surface oxides, as evidenced by cyclic voltammetry (CV) showing significant reduction currents at less negative potentials compared to Cu(100) (Figure 11a). In addition, in situ electrochemical scanning tunneling microscopy (EC-STM) revealed that surface oxidation on Cu(111) enhanced its deoxygenation activity towards ${\mathrm{NO}}_3^{-}$. The Cu₂O formationon Cu surfaces influences their ${\mathrm{NO}}_3^{-}$-to-${\mathrm{NO}}_2^{-}$ conversion efficiency which exhibits higher susceptibility to oxidation and facilitates oxygen transfer from ${\mathrm{NO}}_3^{-}$ and subsequent reduction to ${\mathrm{NO}}_2^{-}$ at less negative potentials. Thus, highlighting Cu₂O role in catalyzing the eNO₃R process [166].

4.1.3. Copper-Based Bimetallic Catalysts

Copper bimetallic catalysts, particularly CuNi alloys and Cu/Ni heterostructures, have shown remarkable potential in electrochemical nitrate reduction due to their synergistic effects and enhanced performance [167]. CuNi alloy cathodes, such as Cu₈₀Ni₂₀ and Cu₅₀Ni₅₀, have demonstrated superior activity in nitrate reduction compared to mono-component metals. The synergistic mechanism between Cu and Ni, where Cu facilitates nitrate reduction while Ni preferentially adsorbs hydrogen atoms, enhances the reduction process by hindering competitive hydrogen adsorption [65,168]. Operando spectroscopy studies have elucidated the structural-performance relationship, revealing the upshifting of Cu d-band center towards the Fermi level upon alloying with Ni. This enhances the adsorption energies of intermediates, reducing overpotential and promoting nitrate reduction [65]. Porous CuNi alloy electrodes, prepared via hydrogen evolution-assisted electrodeposition, exhibit high stability and effective nitrate reduction, attributed to their large surface area and enhanced mass transport [169]. The mechanism proposed by Zhang et al. [84] involves the formation of infinite galvanic cells, facilitating electron transfer and adsorption of nitrate on Cu sites, leading to enhanced reaction kinetics and selectivity (Figure 12). While copper bimetallic catalysts offer promising advantages in nitrate reduction, challenges such as leaching, and particle aggregation need to be addressed for prolonged operation.

4.1.4. Copper Oxide

Copper oxides exhibit significant electrocatalytic activity in the eNO₃R, particularly when engineered with oxygen vacancies (OVs). These vacancies serve as chemisorption sites and facilitate charge transfer, enhancing the catalytic performance of the materials [170]. Daiyan et al. [171] prepared CuO nanoparticles via scalable flame spray pyrolysis (FSP) followed by mild plasma treatment to controllably induce defects. The optimized plasma-treated CuO demonstrated enhanced ${\mathrm{NH}}_4^{+}$ yield (∼292 $\mathsf{\mu }\mathrm{mol}.{\mathrm{h}}^{-1}.{\mathrm{cm}}^{-2}$ at −0.6 V vs. RHE) compared to non-defective CuO. The presence of OVs in CuOx not only strengthens NO₃⁻ adsorption energy but also inhibits the hydrogen evolution reaction (HER), crucial for improving selectivity in the eNO₃R. DFT studies further indicate that most NO₃R intermediates are more stabilized on CuOx with two OVs compared to Cu₂O, highlighting the effectiveness of defect engineering in optimizing catalytic performance [171].

4.1.5. Cu-Based Heterostructure Engineering

Cu-based heterostructures have been extensively studied for their electrocatalytic performance in the conversion of ${\mathrm{NO}}_3^{-}$ to ${\mathrm{NH}}_3$. For instance, Ren et al. [172] designed Cu/Pd/CuOx heterostructures with a porous morphology, which facilitated electron transfer between Cu, Pd, and CuOx components. This heterostructure enhanced ${\mathrm{NH}}_3$ production significantly, achieving a ${\mathrm{NH}}_3$ production rate of 1510.3 μg h⁻¹ mg⁻¹, FE of 86.1%, and ${\mathrm{NH}}_3$ selectivity of 90.06% (Figure 13a). The electronic interactions within the Cu/Pd/CuOx interface promoted the adsorption of intermediates, thus suppressing competitive hydrogen evolution reactions and improving overall catalytic efficiency [172]. Another study by Xu et al. [173] explored Cu₂O corner-etched octahedra supporting Pd nanoparticles (Pd-Cu₂O), where Pd acted as the active center for *H adsorption and catalyzing ${\mathrm{NH}}_3$ production, while Cu₂O provided additional active sites through electrochemical reduction to Cu⁰/Cu⁺. This system exhibited outstanding ${\mathrm{NH}}_3$ formation characteristics, including a ${\mathrm{NH}}_3$ formation rate of 925.11 μg h⁻¹ mg⁻¹, selectivity of 95.31%, and FE of 96.56% (Figure 13b). The oxygen defects in Cu₂O were crucial for enhancing ${\mathrm{NO}}_3^{-}$ adsorption and weakening of the N–O bond, thereby promoting efficient ${\mathrm{NH}}_3$ synthesis [173].

4.1.6. Copper-Based Transition Metal Oxide Composite

Transition metal oxide composite electrodes, particularly those incorporating Cu, have shown promising electrocatalytic activity in nitrate reduction. For example, combining copper oxide (CuO) with cobalt oxide (Co₃O₄) on titanium (Ti) electrodes results in a dramatic increase in nitrate reduction efficiency, reaching nearly 100%. The unoccupied π * orbital and the unclosed d-orbital shells of CuO facilitate nitrate adsorption and its subsequent reduction to nitrites, accelerating the overall nitrate reduction process [174]. Yang et al. [175] proposed a Ti/Cu₅ZnOₓ composite electrode that offers more active sites for the degradation of intermediate nitrogen compounds, especially in alkaline solutions compared to Ti/CuOₓ. The addition of zinc enhances the reduction in nitrite to nitrogen gas (N₂), achieving reduction efficiencies comparable to those of precious metal catalysts like palladium [175]. The process of creating composite electrodes, including those involving copper, is complex and has a low success rate. This complexity limits their large-scale preparation and practical application.

4.1.7. Copper-Based Inorganic/Organic Material

Loading inorganic materials such as graphene and phosphorus onto copper increases the electrode’s stability and specific surface area, which in turn improves the electrocatalytic reduction efficiency of nitrate. Graphene-modified Cu electrodes exhibited a 3.5-fold increase in reduction peak densities compared to pure Cu electrodes, showcasing improved electrocatalytic performance [176]. Polypyrrole (Ppy) coating on copper electrodes enhances the selectivity for ammonia production during nitrate reduction. The mechanism demonstrated that in the low electrode potential range (0.4–0.8 V), Ppy facilitates proton adsorption and electron transfer, which generates atomic hydrogen to convert nitrate to ammonia, improving selectivity and reducing unwanted by-products like nitrite [177]. Additionally, flower-like polycrystalline Cu grown on carbon paper via electrodeposition exhibited remarkable electrochemical activity with an ${\mathrm{NH}}_3$ formation rate of 101.4 μmol h⁻¹ cm⁻² and an FE of 93.91% in neutral solution. The structure facilitated efficient mass transfer of reactants, ensuring sustained performance over multiple cycles [178].

4.1.8. Copper-Based Nanomaterials

Advances in the rational design of micro-/nanostructures have enhanced their efficacy. Notably, copper nanomaterials have gained attraction toward nitrate reduction to ammonia. Cu-nanomaterials used to date are nanosheets, nanowire arrays, nanocubes, nanoparticles, 3D nanobelts, and nanoclusters. Cu-nanosheets (CuNSs) have demonstrated significant efficiency in several studies such as Fu et al. [137] who investigated CuNSs with {111} basal planes, Cu nanocubes ({100}), and irregular Cu nanoparticles. The results indicate that CuNSs displayed exceptional ${\mathrm{NH}}_3$ production selectivity (99.7% FE) at a low overpotential of −0.15 V (vs. RHE). Furthermore, they exhibited distinct reduction peaks in linear sweep voltammetry (LSV), attributed to their specific facets and efficient ${\mathrm{NO}}_2^{-}$ generation, crucial for high eNO₃R performance [137]. Another nanomaterial that has exhibited notable efficiency is nanowire arrays (NWAs), NWAs have been effective due to their reconstructed surface during catalysis, which suppresses HER and achieves high faradaic efficiencies towards ammonia production (up to 95.8%) [58]. Incorporating metals like ruthenium (Ru) on Cu nanowires (Ru-CuNW) has led to exceptionally active catalysts, achieving reduction currents of 1 A cm⁻² at low overpotentials (−0.13 V vs. RHE) and converting over 99% of nitrate into ammonia [179]. Cu-nanobelts have been found to substantially increase the specific surface area, thus boosting the catalytic activity by facilitating better nitrate transfer and accelerating electrocatalytic kinetics. These nanostructured copper electrodes exhibit a high FE of 95.35% and a maximum ${\mathrm{NH}}_3$ production rate of 650 mmol h⁻¹ g⁻¹_cat with a potential 0.15 V vs. RHE, outperforming traditional foam copper electrodes. The enhanced efficiency is attributed to their large surface area and high charge transfer capability, which promotes effective nitrate adsorption and reduction [163]. The advantage of using nanomaterials is their operation at low potentials, which minimizes energy consumption. However, scalability remains a significant barrier to the use of Cu-nanomaterials for nitrate reduction to ammonia.

4.2. Advanced Strategies Toward eNO₃R to Nitrogen

The challenge facing electrochemical nitrate reduction to nitrogen is the low selectivity compared to HER, which means the use of high energy and hinders industrial implementation. In this section, we will discuss the two strategies to advance the process starting with the use of specific materials as in the case of iron-based materials and specifically nanoscale zero-valent iron. Secondly, we address the removal of ammonium by-product generated by electrochemical nitrate reduction with the indirect advanced oxidation with chloride compound.

4.2.1. Nanoscale Zero-Valent Iron (nZVI)

nZVI for nitrate reduction to nitrogen, leveraging various strategies to enhance their catalytic performance. Nanoscale zero-valent iron supported on ordered mesoporous carbon (nZVI/OMC) has demonstrated excellent performance, achieving a maximum nitrate removal capacity of 315 mg N g⁻¹ Fe with a high selectivity towards ${\mathrm{N}}_2$ of up to 74%. The uniform dispersion of iron nanoparticles within the carbon matrix and the large surface area of the electrode contributed to its superior electrocatalytic activity. Additionally, the presence of hydrogen generated during the reduction process not only acted as a reductant but also protected the nZVI from oxidation, thereby enhancing its stability and performance over extended operation periods [143]. Moreover, efforts have been made to explore the mechanism of nitrate reduction mediated by iron-based electrodes, such as polymeric beads supported nZVI (nZVI/D201), which further elucidated the role of in situ formed Fe⁰@FexOy-Fe(II) structures in facilitating the electrochemical reduction process. These advancements underscore the potential of iron-based materials in developing efficient and stable electrodes for nitrate reduction, although challenges related to complex preparation processes remain a consideration for practical applications [144]. Bimetallic nanoparticles, such as iron–nickel (Fe/Ni) and iron-palladium (Fe/Pd), as well as trimetallic nanoparticles like palladium–copper–iron (Pd–Cu–nZVI), represent advanced strategies to boost nitrate reduction efficacy and facilitate the production of non-toxic nitrogen gas. These nanoparticles function synergistically, where the iron component of nZVI acts as an electron donor, initiating nitrate reduction. In bimetallic systems, metals like palladium enhance the conversion of nitrite to nitrogen gas, thereby promoting a more complete detoxification of the water [180]. Furthermore, adjusting the composition ratios, such as maintaining a 2:1 ratio of Pd to Cu in Pd-Cu-NZVI, optimizes the selectivity toward nitrogen gas formation [180]. Durability remains a challenge for nZVI catalysts due to leaching and oxidation. Zhang et al. [181] addressed this by protecting Fe nanoparticles with ultrathin graphene nanosheet layers (Fe@Gnc), maintaining over 96% ${\mathrm{NO}}_3^{-}$ conversion rate and ${\mathrm{N}}_2$ selectivity after 960 h of cycling. This graphene nano chainmail prevents active site agglomeration and accelerates electron/ion transport, suggesting a potential strategy for enhancing the durability and efficiency of Fe-based catalysts for nitrogen conversion processes [181].

4.2.2. Indirect eNO₃R to Nitrogen via Cl-Electrochemical Advanced Oxidation

The ammonia generated during eNO₃R can be oxidized to nitrogen gas by free radicals (${\mathrm{HO}}^{•}$, ${\mathrm{Cl}}^{•}$, ${\mathrm{ClO}}^{•}$) or active chlorine, generated during electrolysis [182]. Studies confirm that ${\mathrm{HO}}^{•}$ alone does not affect ammonia specifically in the absence of chloride [183,184]. In contrast, ${\mathrm{Cl}}^{•}$, ${\mathrm{ClO}}^{•}$, and active chlorine exhibits high ammonia oxidation capacity, achieving nitrogen selectivity exceeding 90% [183,185,186]. Ammonia oxidation in the Cl-EAO process primarily occurs via active chlorine and chlorine radicals. Initially, ${\mathrm{Cl}}^{-}$ ions adsorb onto the anode surface and lose electrons, forming ${\mathrm{Cl}}^{•}$. These radicals may combine to form Cl₂ or react with ${\mathrm{Cl}}^{-}$ to produce Cl₂, which then diffuses into the solution and reacts further to form HClO or ${\mathrm{ClO}}^{-}$. Additionally, ${\mathrm{HO}}^{•}$ is generated on the anode surface, some of which reacts with ${\mathrm{Cl}}^{-}$ in the solution to form ${\mathrm{Cl}}^{•}$. Both ${\mathrm{HO}}^{•}$ and ${\mathrm{Cl}}^{•}$ can further react with HClO or ${\mathrm{ClO}}^{-}$ to form ${\mathrm{ClO}}^{•}$. These species react with ammonia through breakpoint chlorination and chlorine radical oxidation mechanisms, as illustrated in Figure 14 [187]. In the Cl-EAO process, ammonia oxidation occurs at the electrode-solution interface, where the production rate of active chlorine and chlorine radicals is crucial and often limits the overall reaction rate [188]. The rate of chlorine radical production, particularly the production of ${\mathrm{Cl}}^{•}$, is identified as a slower step compared to other reaction rates [189]. During breakpoint chlorination, the mass transfer of active chlorine to water becomes another rate-limiting factor, influencing the rate of chlorine radical oxidation more than breakpoint chlorination itself. At the cathode, active chlorine in the solution is reduced back to ${\mathrm{Cl}}^{-}$. Over time, the concentration of active chlorine in the solution stabilizes, maintaining a stable ${\mathrm{Cl}}^{-}$ concentration due to this reduction process. Despite some Cl₂ escaping from the solution, the impact on active chlorine production remains minimal [187]. Furthermore, the ammonium generated during the process can be influenced by different factors, as demonstrated in

Table 5. Parameters Affecting the Oxidation Process of Ammonium generated.

Influence Factors	Descriptions	Refs.
Initial chloride concentration	high initial concentration of chlorides increases the production of active chlorine and ${\mathrm{Cl}}^{•}$, thereby enhancing Ammonia oxidation	[185,190,191]
pH	pH decreases during the process. Acidic and neutral media are preferable for ammonia oxidation. Alkaline media decrease ammonia oxidation efficiency	[185,192,193]
Current density	Increasing current density improves ammonia oxidation efficiency (rule applied for lower current density than 37.5 mA·cm−2)	[186,192]
Ammonia concentration	High ammonia in solution inhibits the formation of active chlorine and ${\mathrm{Cl}}^{•}$	[194,195]
Temperature	Increasing temperature promotes the oxidation of ammonia. However, at temperatures higher than 55 °C the efficiency decreases due to the lose of active chlorine.	[193]

.

The debatable aspect of this process is the formation of chloramine intermediates, which include NH₂Cl, NHCl₂, and NCl₃, as well as chlorine-containing byproducts (CBPs) such as ${\mathrm{ClO}}_2^{-}, {\mathrm{ClO}}_3^{-}$, and ${\mathrm{ClO}}_4^{-}$ [192,196]. Researchers have found that chloramine concentrations are always below certain specific levels, such as 0.44 mg/L [197]. A study using a graphite/PbO₂ anode showed that the total chloramine concentration was always less than 2.2% of the initial ammonia concentration [192]. Therefore, it is important to focus on the concentration and behavior of chloramines and CBPs in the Cl-EAO process to enhance the understanding and application for the removal of ammonium generated during the electrochemical reduction in nitrates.

5. Application of eNO₃R to Brackish Groundwater: A Case Study

The groundwater is typified by elevated levels of chloride, potassium, sodium, magnesium, calcium, sulfate, bicarbonate, and carbonate, culminating in the formation of a brackish environment. Furthermore, it is important to acknowledge that the quality of groundwater is subject to various influences, including environmental and geological factors such as residence periods, climate conditions, soil properties, flow patterns, terrains, the chemistry of the replenishment zone, local geology, and water intermixing [198,199]. Additionally, nitrate derived from fertilizers easily leaches through soils and enters groundwater, especially following periods of heavy rainfall. The application of electrochemical processes remains the most rational and applicable compared to other technologies. However, before applying this method to large-scale water treatment, it is essential to discuss the parameters that significantly affect the nitrate reduction process, including pH, nitrate concentration, coexisting ions, current density, and applied voltage analyze the water matrix, and identify potential issues that may arise.

5.1. pH

During the reduction process, electrocatalytic nitrate reduction consumes protons in both the rate-limiting step and the subsequent reduction steps, which means that the pH increases with the electrolysis process. A thorough investigation was carried out to identify the optimal circumstances in which the pH could exert an influence. eNO₃R process indicated that the best removal rates were achieved under acidic conditions, in contrast to neutral conditions that led to poor results [105]. Lowering the pH of the electrolyte solution, used in electrochemical nitrate reduction processes, can also provide additional benefits as reported by Yao et al. [200] where an electrolyte with a lower pH facilitates the formation of ions (e.g., ${\mathrm{Ti}}^{2+}\mathrm{or}/{\mathrm{and} \mathrm{Ti}}^{3+}$), which can reduce nitrates and nitrites. On the other hand, the Pourbaix diagram of the ${\mathrm{N}}_2–{\mathrm{H}}_2\mathrm{O}$ system (Figure 15) indicated that the hydrogen evolution reaction (HER) cannot be ignored as it competes with the eNO₃R due to electron consumption. The FE of HER is higher in acidic media than in alkaline media, which results in a significantly lower FE of nitrate reduction in acidic media compared to alkaline media. Due to competition with ${\mathrm{H}}_2$, the direct charge transfer process has low efficiency. However, the indirect electrochemical autocatalytic process (Vetter and/or Schmid) can be improved under acidic conditions due to the compensation of this effect. Zhang et al. [181] found that a neutral pH of 7, was the most favorable for electrocatalytic nitrate reduction, as acidic conditions caused hydrogen bubbles to form and prevent contact between the active site and the electrolyte, while higher pH led to the active surface being occupied by adsorbed hydroxide. On the other hand, Beltrame et al. [201] indicated that solutions with pH values between 6.0 and 6.5 induced a lower conversion of nitrate than in alkaline media. Notably, nitrite formation was significantly minimized, and gaseous compound generation was higher in solutions at given pH values due to the availability of ${\mathrm{H}}_{\left(\mathrm{aq}\right)}^{+}$, which in turn promotes gaseous compound generation.

5.2. Effect of Chloride Anions

The primary ions present are chlorides, which are found in substantial concentrations in this type of water. Generally, the pathway for ammonium production by eNO₃R is limited in this case, as it necessitates a cathode with high faradaic efficiency (FE) and an anode with the highest oxygen evolution reaction (OER) compared to chlorine ion oxidation reactions (ClOR). Additionally, research on this topic is currently limited. Thermodynamically, ClOR has higher potentials compared to OER, making it prone to activation under high cell voltages required for industrial current densities. In acidic conditions, ClOR manifests as a chlorine evolution reaction (ClER), while in alkaline conditions, it forms hypochlorite, both of which are corrosive and detrimental to catalyst longevity [203]. Materials such as Mn-based oxides and non-noble metal catalysts like transition metal hexacyanometallates [204], cobalt phosphates [205], selenides [206], and borates [207] have shown promise in stabilizing OER in NaCl-containing electrolytes. However, they struggle to deliver industrially relevant current densities at overpotentials below 490 mV. Recently, electrocatalysts resistant to chloride ion oxidation for industrially relevant current densities and low overpotentials have been extensively studied, even though their primary application is aimed at hydrogen production from seawater. Researchers such as Dong et al. [208] have developed a robust electrocatalyst based on nickel–iron hydroxide with low nucleophilicity. The EA-Ni_0.82Fe_0.18-OH electrocatalyst provides industrially relevant current densities of 400 and 1200 mA cm⁻² in simulated alkaline seawater, with respective overpotentials of 312 mV and 411 mV. Moreover, it exhibits impressive durability at current densities of 500 mA cm⁻² over 50 h of simulated alkaline seawater electrolysis [208]. We need to advance research in developing anodes that will play a crucial role in the large-scale application of ammonium production from brackish waters. On the other hand, the pathway for nitrogen selectivity remains reasonable and motivates its large-scale application, especially with the presence of a high chloride concentration in this type of water, which plays a significant role in converting generated ammonium to gaseous nitrogen, as discussed in the previous section (Cl-electrochemical advanced oxidation).

5.3. Coexistence Ions

The influence of the nature of the coexisting ions has been largely reported. For instance, Katsuunaros et al. [122] noted that the reduction in nitrate increases slightly with an increase in NaCl concentration. Additionally, according to the ${\mathrm{Li}}^{+}<{\mathrm{Na}}^{+}<{\mathrm{K}}^{+}<{\mathrm{Cs}}^{+}$series of the supporting electrolyte, the reduction rate also increases significantly. Likewise, they demonstrated that the rate of reduction is notably higher in the presence of multivalent cations, namely ${\mathrm{Ca}}^{2+}$ and ${\mathrm{La}}^{3+}$ as well as ${\mathrm{NH}}_4^{+}$ [122]. Moreover, Shen et al. [209] showed that when the concentration of ${\mathrm{Na}}_2{\mathrm{SO}}_4$ was high, the rate of removal of ${\mathrm{NO}}_3^{-}$ was rapid and this can be explained by the fact that a higher concentration of electrolytes induced a higher rate of electron transfer and the effect of electrolytes on the selectivity to ${\mathrm{N}}_2$ was remarkable. The effect of the supporting electrolyte cation on nitrate reduction has been explained by the theory of “cationic catalysis”, which suggests that the cation of the supporting electrolyte acts as an attracting center for the reduced anion by forming an instantly neutral ion pair. The latter is not repelled by the negatively charged electrode, increasing the rate of nitrate reduction [122]. Manzo-Robledo et al. [210] evaluated the effect of ${\mathrm{K}}^{+}$ and ${\mathrm{Na}}^{+}$ cations on nitrate reduction in an electrolyte solution using real-time on-line differential electrochemical mass spectrometry. The results indicated that the presence of ${\mathrm{K}}^{+}$ led to the production of ${\mathrm{N}}_2$ as the main product, while ${\mathrm{Na}}^{+}$ favored the production of ${\mathrm{N}}_2\mathrm{O}$. The study also reported a moderate HER in ${\mathrm{K}}^{+}$-containing electrolyte solutions, promoting the reaction between ${\mathrm{NO}}_3^{-}$ and ${\mathrm{H}}_2$ species, while the HER kinetics was more dominant in ${\mathrm{Na}}^{+}$-containing electrolyte solutions, leading to a suppression of molecular nitrogen generation. These results highlight the significant influence of different cations in the supporting electrolyte on the products and kinetics of nitrate reduction [210]. Fedurco et al. [211] conducted also a study to investigate the influence of ${\mathrm{La}}^{3+}$ and ${\mathrm{Al}}^{3+}$ cations on nitrate reduction at a polycrystalline silver electrode. The outcomes showed that the presence of ${\mathrm{La}}^{3+}$ caused a positive shift in the onset potential, indicating an enhancement of nitrate reduction. This may be attributed to acid-base catalysis, with La(III) species acting as proton donors and/or OH⁻ ion acceptors at the electrode surface. The research studies also exhibited that films deposited onto the electrode surface consisted of $\mathrm{La}{(\mathrm{OH})}_3$ and/or $\mathrm{Al}{(\mathrm{OH})}_3$ in the solutions containing ${\mathrm{La}}^{3+}$ and ${\mathrm{Al}}^{3+}$ ions, respectively. The La(III) species in the solution then diffused to the electrode surface as nitrate complexes or ion pairs, thereby promoting the reduction reaction. In contrast, ${\mathrm{Al}}^{3+}$ cations, which do not form ion pairs with ${\mathrm{NO}}_3^{-}$ anions were found to be less effective in promoting the nitrate reduction reaction [211]. In contrast, the rate of reduction was reduced by the anion of the supporting electrolyte. Specifically, ${\mathrm{I}}^{-}$ exhibited the greatest inhibitory effect, followed by ${\mathrm{Br}}^{-}$, ${\mathrm{Cl}}^{-}$, and ${\mathrm{F}}^{-}$. Beltrame et al. [201] noticed that pH control by sulfuric acid affected nitrate reduction, as the ${\mathrm{SO}}_4^{2-}$ anion present in the solution can interfere with nitrate adsorption in the active sites of the catalyst, leading to a diminution of nitrate reduction.

5.4. Cathode/Anode Passivation

The presence of calcium, magnesium, and carbonates within brackish groundwater constitutes a significant impediment to the development of a large-scale process. In high pH environments, cations like Ca²⁺ and Mg²⁺ can contribute to the formation of a passivating surface layer on the cathode duo to electrophoretic migration phenomena. Furthermore, calcium and magnesium readily react with hydroxides formed during electrolysis and carbonates to produce insoluble compounds that form a layer that deposits on the cathode surface, thereby reducing the efficiency of nitrate reduction to nitrogen and ammonium [212]. Indeed, among the strategies proposed to reduce cathode passivation is using pulsed current instead of direct current [213]. So far, there have been few studies aimed at investigating the effect of pulsed current such as Li et al. [101] who demonstrated that the use of pulsed electrolysis (PE) effectively inhibited the HER across all PE experiments compared to constant potential electrolysis (CE), accompanied by lower energy consumption. Additionally, Cu single-atom-modified gels (Cu SAGs) used in the study also exhibited good stability after ten successive catalytic cycles, with a slight performance decrease noted after the seventh cycle. Therefore, it is crucial to focus on this topic in future research concerning this issue, while also emphasizing the stability of the cathode [101]. Now the question arises: Is this strategy advantageous for the anode? The clearest aspect is that pulsed currents can momentarily disrupt the accumulation of ions at the electrode surface, which contributes to delaying the decay of the anode potential. However, the question of whether the application of dual-pulsed or multi-pulsed electro-oxidation is more beneficial has already been addressed in the research by Jiang et al. [214] in which they showed that Multi-Pulse Electro-Oxidation has the highest total nitrogen removal compared to Direct Current Electro-Oxidation and Dual-Pulse Electro-Oxidation. This mode of MP-EO perfectly improves the mass transfer and specifically during the T_off period between pulses, the oxidants and intermediates produced can efficiently transfer into other areas, thereby improving the overall efficiency of the process. However, the dual-pulsed approach, which was expected to yield favorable results, showed the opposite in this study. This outcome is attributed to the type of anode used, specifically stainless steel. It is evident that this material is not designed for the specific chlorine evolution reactions required [214]. Other specific types of materials, such as Ti/Ru-Ir, have not yet been tested in this context and for the same specific conditions to determine whether the dual-pulsed effect on ammonium removal efficiency is truly superior to that of direct current.

5.5. Nitrate Concentration

The variation in nitrate concentration also caused changes in the kinetic order of the electrode during electrochemical reduction. In acidic solutions, Dima et al. [39] studied the electrocatalytic reduction in nitrate at low concentrations using noble and transition-metal electrodes. They found that cathodic currents increased as the concentration of nitrate augmented. The Tafel slopes indicated that the rate-determining step involved the first electron transfer and was close to or greater than 120 mV. dec⁻¹. The nitrate reaction order was found to be approximately 0.51 for Pt and 0.34 for Rh from the logarithm linear fitting of the current densities versus the logarithm of nitrate concentration [39]. Nitrate adsorbates occupied the active surface area, causing the butterfly-like hydrogen adsorption and desorption pattern to disappear. Also, Taguchi et al. [215] assessed the impact of nitrate concentration on the cyclic voltammogram of Pt(110) in 0.1 M ${\mathrm{HClO}}_4$ + 10 mM ${\mathrm{KNO}}_3$, and concluded that the reaction follows a zero-order dependence on nitrate concentration. The adsorbed ${\mathrm{NO}}_3^{-}$ is believed to be in a state of quasi-equilibrium with the ${\mathrm{NO}}_3^{-}$ in the bulk solution before the rate-determining step of the reduction reaction. The average Tafel slope value of −66 ± 2 mV. dec⁻¹ indicating that the RDS does not involve any electron transfer. Instead, a pure chemical reaction occurs on Pt(100), wherein ${\mathrm{NO}}_3^{-}$ reacts with ${\mathrm{H}}_{\mathrm{ads}}$ [215]. In another study, Katsounaros et al. [122] studied the kinetics of nitrate reduction at high concentrations and determined that the reduction follows Langmuir–Hinshelwood kinetics. Whereas the reduction is characterized by first-order kinetics when the concentration is below 0.3 M. However, at higher concentrations, zero-order kinetics dominate due to the blockage of surface sites, indicating that the regeneration of active sites controls the reaction. Also, the selectivity towards ${\mathrm{N}}_2$ increased from 70% to 83% with an increase in nitrate concentration from 100 to 1500 mg L⁻¹ and remained relatively stable at higher nitrate concentrations. Conversely, the selectivity of ammonia showed a declining trend from 25% to 11%. Furthermore, the faradic efficiency increased from 25% to 78% with an increase in nitrate concentration, and this trend was observed when 95% of nitrate was reduced [122].

5.6. Influence of Applied Voltage and Current Density

The applied voltage or potential had a relatively strong effect on nitrate reduction. Ding et al. [216] demonstrated that the selectivity of nitrate reduction in products is strongly dependent on the potential applied. In addition, Reyter et al. [217] showed that at a potential of −0.9 V, nitrite is the only substance generated at the cathode, which is then oxidized at the anode, thus decreasing the process efficiency [217]. Also, Shen et al. [209] found that the application of increasingly negative cathodic potentials resulted in an improvement in nitrate removal efficiency. Specifically, a more negative potential facilitated the rapid removal of nitrate due to the thermodynamically favorable reduction in nitrite to ${\mathrm{N}}_2$ occurring at more negative potentials. On the other hand, Zhao et al. [218] showed a decrease in selectivity for ${\mathrm{N}}_2$ due to the reduction in ${\mathrm{Cu}}_2\mathrm{O}$ to metallic Cu caused by the application of a high cathodic potential. It is worth noting that the application of very negative potentials may lead to the formation of undesirable by-products such as ammonia or hydrogen gas. Hence, the cathodic potential should be optimized to achieve both high nitrate removal efficiency and high selectivity for nitrogen gas.

In addition to the cathodic potential, the current density also plays a significant role in the performance of nitrate removal. In this regard, Yao et al. [200] demonstrated that nitrate removal efficiency increased progressively with increasing current density. It also caused an acceleration of electrode corrosion and produced reducing agents that enhanced nitrate reduction. Gao et al. [219] found that as the current density increased, more energy was utilized to produce hydrogen bubbles. This, in turn, resulted in a reduced selectivity of gaseous nitrogen and an increased selectivity of ammonia, ultimately leading to a lower overall rate of total nitrogen removal.

5.7. Challenge to Move from Lab Scale to Pilot eNO₃R

The eNO₃R reactors in most of all previous research do not exceed 500 mL. However, large-scale application requires several factors to ensure successful applicability, specifically for brackish groundwater. One of the initial considerations in designing a reactor is selecting between divided and undivided reactors. For nitrate removal from brackish groundwater, the choice between a single chamber cell (SCC) and a double chamber cell (DCC) is directly linked to the objective. If the goal is to convert nitrates into gaseous nitrogen, the choice should be an Undivided Flow electrochemical reactor (UFECR). However, if the objective is ammonium production and an inert anode for chlorides is used, as discussed in the previous section, the preference remains for UFECRs. On the other hand, if an anode with higher ClOR is used, then the choice is a Divided Flow Electrochemical reactor (DFECR), as explained in Figure 16.

Studies employing continuous mode which allows for easier scale-up and industrial application are few. Typical continuous reactors are illustrated in Figure 17. For example, Abdulah et al. [220] used a homemade flow cell with a porous modified copper cathode and platinum anode to convert nitrate to ammonium; the results showed a maximum ${\mathrm{NH}}_4^{+}$ selectivity of 96% and a high faradaic efficiency for ${\mathrm{NH}}_4^{+}$ formation (FE 72%) was recorded with a solution of pH 7.2, and a flow rate of 2 mL min⁻¹ [220]. Even though the water matrix used differs from that of brackish groundwater, the results remain significant and encourage the development of DFECRs for ammonia production. Another study by Makover et al. [221] demonstrated significant results regarding the use of continuous mode; however, the application of periodic current reversal complicates the use of these processes for large-scale applications, particularly for natural waters, due to cathode dissolution [221]. Another study utilized a 500-L pilot continuous-flow reactor for nitrate reduction to ammonia [222]. The stable operation achieved a ${\mathrm{NH}}_4^{+}$–N selectivity of 92.9% at a hydraulic retention time (HRT) of 12 h and a current density of 20 mA cm⁻². However, the long hydraulic retention time required for a significant decrease in the initial ${\mathrm{NO}}_3^{-}$–N concentration is one of the weaknesses of this operation. This is due to non-uniform flow and the presence of numerous dead zones in the reactors, demonstrating that using a continuous-flow reactor for nitrates is not always straightforward [222]. However, the major challenge is transitioning to continuous mode with good hydraulic flow and minimal dead zones to ensure uniform nitrate reduction. Therefore, selecting the appropriate geometry and operating conditions for electrochemical reactors (ECRs) to achieve high current efficiencies (close to 100%) toward nitrate reduction to nitrogen or ammonium in brackish groundwater, requires intensive research into transport phenomena such as hydrodynamics, potential-current distributions, heat, and mass transfer. Mathematical modeling and Computational Fluid Dynamics (CFD) simulations have facilitated the design, characterization, and scaling up of ECRs. Fluids dynamics phenomena, fluid pressure drop, and current-potential distribution for ECRs have been widely studied using CFD. However, mass transfer phenomenon is an important aspect to be analyzed with CFD for eNO₃R [223]. Many mass transfer issues in electrochemical reactors have been solved by neglecting the migration term, which simplifies practical assumptions, as in a recent study by Oriol et al. [224] who utilized a rotating cylinder electrode reactor (RCE) for nitrate reduction. The results indicate that 90% nitrate was removed in 10 min for a solution comprising 10 mM NO₃⁻ and 500 mM K₂SO₄, at 1000 rpm and 25 °C confirming the RCE reactor’s effective performance for nitrate reduction. Furthermore, CFD showed that mass transport in RCE reactor is linked to the hydrodynamic pattern, highlighting the critical role of fluid dynamics in reactors, particularly in nitrate reduction, to ensure effective mass transfer. Therefore, thorough studies using CFD are necessary to design future types of electrochemical reactors [224]. Another strategy was recently innovated by Zhou et al. [225] to address the mass transfer issue for nitrate reduction, utilizing a flow-through zero-gap electrochemical (ZGEC) reactor. The system demonstrated significant success, achieving 100% nitrate conversion and a rapid reduction kinetics of 0.07676 min⁻¹ [225]. These two studies confirm that focusing on mass transfer is crucial to ensure the success of eNO₃R for large-scale applications.

5.8. Economic Analysis

Most publications in nitrate removal by electrocatalysis focus on the performance of electrode material but rarely discuss capital expenditures (CAPEX) and operational expenditures (OPEX). However, it is obvious that since the process is still in the R&D phase, significant steps remain before industrial application, such as the construction of efficient reactors [226]. Nonetheless, this does not dismiss the importance of making estimates to determine whether the process is economically viable or not. Generally, 90% of synthetic ammonia is produced using the Haber–Bosch process, but there are other methods such as electrocatalytic and photocatalytic ${\mathrm{N}}_2$ reduction. Therefore, eNO₃R remains innovative compared to existing methods [227,228]. The comparison between energy consumption and energy source for the different processes shows that eNO₃R is the most promising due to the total energy demand during the process. Furthermore, the possibility of using solar energy remains a strong point of the process. The production cost of ammonia by the Haber–Bosch process is $560 per metric ton, which is close to the estimated value for production by eNO₃R in the case of using the strained ruthenium nanocluster ($776 per metric ton) [229]. Additionally, Joshua M. McEnaney et al. [230] conducted an economic study based on NH₄NO₃ as the final product instead of ${\mathrm{NH}}_3$. They validated the production of this type of fertilizer at a cost that matches or is lower than the typical United States Department of Agriculture (USDA) cost range for NH₄NO₃. Nonetheless, strictly with low electricity costs and reasonable Faradaic efficiencies (Figure 18) [230]. On the other hand, the techno-economic analysis by Jianan et al. [231] confirmed the relationship between the energy cost per kilogram of (NH₄)₂SO₄ and the Faradaic efficiency/energy consumption. They also emphasized that the concentration of nitrates in effluents plays a crucial role in minimizing costs. So, a high concentration of 1.0 M nitrate found in wastewater is preferable to 0.01 M found in some brackish groundwater to minimize the cost of ammonia production [231]. Overall, eNO₃R for ammonia or even fertilizers production has potential economic benefits and is a promising candidate to replace the Haber–Bosch process specifically for high concentrated nitrate contaminated water. Regarding the concentration of nitrate in brackish groundwater, it is preferable to convert it to gaseous nitrogen, which is economically viable. However, the cost of electrode materials and energy consumption are critical to the application of this process. So far, there have been no economic analyses of this process because it is still in the stage of researching selective materials. However, it is essential to consider the cost of the materials used in the process, focusing on non-noble or non-metallic materials in future research to compete with existing processes.

5.9. Sustainability Analysis

As previously discussed, electrocatalytic nitrate reduction (eNO₃R) processes are gaining attraction as alternative processes for the removal of nitrate from various water effluents, including wastewater and groundwater [232]. Indeed, the main reason for this is that these techniques are regarded as environmentally friendly and efficient approaches. Nonetheless, the main stumbling blocks to implementing these technologies on an industrial scale are energy consumption and cost [232,233,234].

Thus, renewable energies (RE), particularly solar energy, are seen as an alternative to conventional electricity, as they could convert these processes into truly sustainable, eco-friendly, and autonomous technologies, as well as lessen the environmental impacts related to fossil energy [232,235]. In fact, solar energy can be mobilized to power the electrochemical cells employed in nitrate reduction, which on the one hand will lead to an increase in the NH₃ faradaic efficiency, and on the other hand, reduce reliance on non-renewable energy sources and subsequently decrease the overall carbon footprint. In this context, Kani et al. [236] harnessed solar energy for the ${\mathrm{NO}}_3^{-}$ electroreduction to NH₃ using oxide-derived Co as the catalyst. As a result, they attained a high NH₃ faradaic efficiency of 92.37 ± 6.7% current density of 565.26 mA.cm⁻² at −0.8 V vs. RHE, and a high solar-to-fuel efficiency of 11%. Furthermore, hybrid energy systems that couple solar power with other renewable sources like wind or hydroelectric power could guarantee a more consistent and reliable energy source, thus overcoming the intermittency problems inherent in solar power alone. Additionally, energy storage solutions including batteries or supercapacitors can also boost system stability and effectiveness, maintaining continuous operation during times when solar power is unavailable [232].

The integration of solar and wind RE enhances the environmental and sustainability aspects of electrochemical processes but can also have adverse environmental impacts. For instance, water and land use, habitat loss, and the use of harmful materials in manufacturing can be viewed as potential negative effects in the case of photovoltaic (PV) solar panels [237]. Similarly, the deployment of wind turbines can lead to the depletion or degradation of habitat for wildlife, fish, and plants. Rotating turbine blades can also pose a threat to flying wildlife, and cause “noise pollution” [238]. Whilst many studies have documented these environmental impacts, it remains the case that renewable energies contribute to reducing conventional energy consumption and carbon emissions, as well as supporting local development [239,240]. It is, therefore, vital to carry out life cycle assessment (LCA) studies to list and evaluate all potential environmental impacts, so as to be able to draw up an in-depth action plan to reduce the harmful impacts listed above and take advantage of the positive points of implementing renewable energies.

In parallel, a Life Cycle Assessment of electrocatalytic nitrate reduction is essential to fully assess its environmental sustainability. LCA evaluates the environmental emissions and impacts on human health associated with the entire life cycle of the process, from raw material extraction to final disposal [241,242]. This analysis must consider factors such as energy consumption, emissions, and waste production. When comparing these impacts with those of conventional nitrate removal processes, i.e., ion exchange, reverse osmosis, or electrodialysis processes, we can better identify the relative advantages and limitations of electrocatalytic reduction. Furthermore, key sustainability metrics, such as carbon footprint and water use, can provide quantitative information on the environmental performance of the process and emphasize areas for enhancement and optimization. To the best of the authors’ knowledge, there are no previous literature studies on the LCA of electrocatalytic nitrate reduction for brackish groundwater treatment. However, Mosalpuri et al. [243] carried out a techno-economic analysis and an LCA of the electrochemical conversion of nitrate ions present in wastewater to hydroxylamine (NH₂OH), which is a valuable chemical intermediate. They calculated the NH₂OH production costs and determined the life cycle emissions for a small-scale facility (producing 1500 kgNH₂OH/day) and a large-scale facility (producing 50,000 kgNH₂OH/day) integrated into an Iowa wastewater treatment plant. Thus, they found that the NH₂OH production costs for the small- and large-scale facilities are estimated at $6.14/kg-NH₂OH and $5.37/kg-NH₂OH, respectively. The parameters dominating the electrochemical reactor cost are electrolyte, separations, and fixed cost. Meanwhile, LCA results revealed that the proposed electrochemical pathway to produce NH₂OH has lower life cycle impacts than the conventional pathway, especially when using solar energy. Bearing in mind that the Haber–Bosch process entails significant environmental impacts, such as high energy consumption and high CO₂ emissions (400 t of CO₂ per year, or 1.6% of global CO₂ emissions), mostly due to the production of H₂ by steam methane reforming (SMR), which releases CO₂ stoichiometrically [244,245].

In light of what has been described above, it can be deduced that a great deal of research is still required, both at the laboratory and pilot scale, to bring electrocatalytic processes to a high level of technological maturity, but they are also attractive in terms of prospects and applications. Accordingly, there are both challenges and opportunities in implementing renewable and sustainable energy solutions in electrocatalytic nitrate reduction processes. The major challenge is to scale up these systems to the industrial scale, whilst keeping economic feasibility by considering the initial costs of renewable energy infrastructure and storage technologies, which can be significant. Technological innovations, including advances in electrode materials, catalysts, and energy conversion systems, are key to enhancing efficiency, reducing costs, and making sustainable solutions economically viable. In the future, the integration of emerging trends in renewable energy and environmental regulations is likely to shape the future of sustainable nitrate reduction technologies, offering both opportunities for progress and challenges.

6. Summary and Outlook

This review provides an overview of the electrocatalytic nitrate reduction (eNO₃R) to nitrogen and ammonia for application to brackish groundwater. A comprehensive analysis of the eNO₃R mechanism and the theoretical simulations by DFT were presented to predict the reaction pathway of electrocatalysts for ${\mathrm{NH}}_3$ production and ${\mathrm{N}}_2$ selectivity. An overview of copper-based materials for eNO₃R to ammonia as efficient transition metal-based catalysts is given in terms of single-atom copper catalysts, facet engineering, bimetallic, Cu oxide, Cu transition metal oxide composite, Cu-based inorganic/organic material, and Cu-based nanomaterials. Strategies to convert nitrate to nitrogen with iron-based catalysts specifically nanoscale zero-valent iron (nZVI) and strategies to convert ammonia generated with chlorine radical and active chlorine are discussed. A comprehensive analysis of eNO₃R for brackish groundwater as a case study with the various operational parameters that influence the process with the challenge of moving towards large-scale application alongside an economic analysis was reported. This comprehensive analysis is essential for researchers and practitioners, as it provides critical information for the design and optimization of nitrate removal systems for this kind of natural water. eNO₃R holds great promise for removing nitrate from contaminated water sources, including brackish groundwater, but it should be noted that this technology is still in its infancy and several challenges must be addressed. The following recommendations and information could be considered for the design and optimization of an efficient electrochemical cell unit and the selection of the preferred pathway for this type of water:

• The mechanism for the eNO₃R process is not well understood, which could be a reason to continue using DFT calculations. Farther, the complexity of the nitrate reduction process may make it difficult to accurately model using DFT. On the other hand, the DFT calculations require significant computational resources, which can be time-consuming and expensive. Additionally, DFT calculations rely on several approximations and assumptions, which may not always accurately reflect the real-world system. Despite these challenges, the use of DFT and microkinetic modeling can still provide valuable insights into the mechanism of electrocatalytic nitrate reduction and guide the tailored design of more efficient and selective electrode materials.
• Machine learning (ML) and artificial intelligence (AI) play an increasingly important role in optimizing catalyst design, by enhancing the efficiency of DFT. These tools reduce the complexity of high-dimensional catalytic activity maps by identifying key descriptors, such as adsorption energies of critical species. This approach allows for more efficient screening of potential catalysts across vast chemical spaces, accelerating material discovery while significantly lowering the computational cost of DFT calculations.
• Developing more efficient electrocatalysts for eNO₃R is needed to find non-precious metal electrode materials that are highly efficient and durable, as the scarcity and cost of noble metals like Pd make them unsuitable for practical use. This requires extensive research and testing of various materials and modifications to improve their stability and durability, as well as their selectivity and activity for electrochemical nitrate reduction. Furthermore, advanced research in nanomaterials for Cu-nanomaterials or Fe-nanomaterials is required due to their operational low potential which reduces energy consumption for the process.
• Brackish groundwater with high chloride concentration can be advantageous for nitrate conversion to nitrogen; an efficient anode of Ti/Ru-Ir is widely used for this purpose. However, by-product formation such as chlorine-containing byproducts (CBPs) can be the primary drawback, especially with water-containing organic compounds. Focusing on a future hybrid process that includes photo-electrocatalysis or electrocoagulation may be an effective strategy for addressing this special issue.
• Optimizing operational parameters such as pH, current density, and applied voltage is crucial to improve eNO₃R and their effects should be examined in real environments as in the case of brackish groundwater. Additionally, further studies are needed, especially on the challenge of anode and cathode passivation which results in fouling of the electrodes and seriously decrease the efficiency of nitrate reduction, notably on the application of electric pulses and the specific type to ensure the performance of the process and prevent metal leaching. This can enhance nitrate removal efficiency, minimize by-product formation, and promote a sustainable process.
• Selecting the configuration type for the cell, divided or undivided, can greatly impact the efficiency and selectivity of the reduction process. Divided cells are more effective at nitrate reduction, and they could be the suitable choice for ammonia generation as they prevent re-oxidation of the reduced nitrate in the anode and concurrently, we recover the ammonium that migrates through the cation exchange membrane (CEM). However, the drawbacks associated with divided cells are fouling and scaling of the membrane in complex water matrices as the case in brackish groundwater, additionally, the limited availability of ammonium-selective membranes needs to be addressed in future research. The alternative pathway for eNO₃R to ammonium in brackish groundwater can be employed in undivided cells, providing a specific anode that prevents the production of chlorine radicals and active chlorine. These anodes are currently being developed, especially for large-scale applications. On the other hand, undivided cells remain the optimal choice for eNO₃R to ${\mathrm{N}}_2$ due to their effectiveness in minimizing energy consumption and operational costs, while also simplifying system design and maintenance.
• Reactor design and the components of the cells, such as shapes, dimensions of electrodes, and their placement in the flow channel are essential for the success of the process due to their role in ensuring fluid dispersion, current distribution, and mass transport. Mathematical approaches to designing electrochemical reactors have been adopted for different technology using numerical tools such as computational fluids dynamics (CFD) since we can test the various reactors with different components before the experimental characterization in laboratory-scale cells and then, during the scale-up of pilot plants. Therefore, the use of CFD simulation to visualize surface-phase changes during electrode reactions, including the mass transport and current distribution during eNO₃R is crucial to increase the knowledge and performance of the process. Furthermore, designing a new reactor to lower the ohmic resistance and thus reduce unnecessary energy consumption is also a desirable approach.
• Researchers who focus on the conversion of eNO₃R for ammonia production to pursue subsequent utilization should consider the concentration of nitrate in water. Using renewable energies for cost minimization remains advantageous. However, in waters with low nitrate concentrations, it is preferable to prioritize the production of drinking water, specifically through the pathway of converting nitrates into nitrogen. Nevertheless, the generation of ammonium from chemical industry waste, sewage, and the site of caliche ore with high nitrate concentration is the most effective approach to address “two birds in one stone”, removing nitrate and producing ammonia.
• The operating cost requires special consideration for large-scale applications, especially raw materials, transportation, and post-treatment costs. Also, the electrical energy consumed during nitrate electroreduction represents a large proportion of the total cost. In fact, coupling thermodynamically favorable reactions such as the oxidation of organics is promising, which can not only decrease the electrical energy needed but can also raise the additional value of electrochemical water treatment.
• The inclusion of renewable energies, notably solar power, in electrocatalytic nitrate reduction (eNO₃R) promotes sustainable development by minimizing dependence on non-renewable sources and minimizing the carbon footprint. Moreover, the use of Life Cycle Assessment (LCA) is essential to assess the environmental impact of eNO₃R compared to conventional methods, concentrating on energy consumption, emissions, and waste. Yet in-depth research studies are called for to address challenges such as scale-up, high initial costs, and the requirement for technological innovation for sustainable implementation, with the potential support of economic incentives and emerging renewable energy trends.