Rational design on photo(electro)catalysts for artificial nitrogen looping

Nitrogen, one of most important elements on the Earth, plays an essential role in shaping the modern society. The natural nitrogen looping, however, is insufficient to satisfy the high demand of the large-scale human activities. To achieve a more sustainable and efficient utilization of nitrogen, artificial nitrogen


| INTRODUCTION
As one of the most abundant elements on Earth, nitrogen plays a key role on the daily life and industry in modern society. For example, nitrogen-based fertilizers (eg, ammonium sulfate, ammonium nitrate, ammonium bicarbonate, urea, etc.) are the frequently used artificial chemical fertilizers in agriculture, which provides the sufficient food supply for the rapidly growing world population. 1,2 Besides, various nitrogenous chemicals, such as ammonia (NH 3 ), amine, and cyanogen, are also the important primary chemical products or clean energy carriers. Specifically, NH 3 is not only an essential raw material but also an ideal carrier of renewable energy (eg, ammonia-based fuel cells and hydrogen storage medium) 3,4 ; while amine has the significant application in dye industry and drug production. As for cyanogen, it is available for industrial electroplating technology. 5,6 As the major resource of elemental nitrogen, nitrogen gas (N 2 ) accounts for approximately 78% of the air in volume. Therefore, its reserves are considered very abundant, which enables the possible large-scale utilization of nitrogen. Unfortunately, the N≡N triple-bond in such dinitrogen molecules (ie, N 2 ) exhibits an extremely high binding energy of approximately 941 kJ/mol, leading to a significant inertness for activating nitrogen gas. Consequently, it is very difficult for the direct utilization of the nitrogen gas as a raw material to produce various nitrogen-based chemicals. [7][8][9][10] Obviously, the nitrogen fixation process that opens the strong N≡N triple-bond to convert nitrogen element from the free state to the combined state has become a pivotal step for the effective utilization of nitrogen element. Typically, nitrogen fixation process can be categorized into two groups: the natural and artificial routes. To be more specific, the natural nitrogen fixation refers to the nitrogen reduction to ammonia by nitrogen-fixing enzymes that are composed of ferritin and molybdoferredoxin in rhizobia. 11,12 This process is the principal approach for plants to obtain nitrogen element, and it is one of the most fundamental reactions in the food chain in nature as well. However, the nitrogen fixation process in nature exhibits an extremely low efficiency, which cannot satisfy the current increasing demand for chemical fertilizers and nitrogenous chemicals. Consequently, exploration for more efficient artificial routes for nitrogen fixation is not only necessary but also urgent. [13][14][15][16] As early as the 19th century, some researchers proposed the necessity of artificial nitrogen fixation for meeting the growing need for food. It is not until the early 20th century, however, that Fritz Haber and Carl Bosch developed the efficient method of artificial nitrogen fixation, namely the Haber-Bosch process. In this catalytic process, the conversion from nitrogen gas to ammonia was achieved under the high temperature (300 C-500 C) and the high pressure (200-300 atm), simultaneously assisted by the Fe/Ru-based catalysts. The Haber-Bosch process achieved the large-scale implementation of industrial ammonia synthesis, which laid a solid foundation for the development of modern society.
The aforementioned Haber-Bosch process, nevertheless, has been regarded as an unsustainable strategy due to its huge energy consumption and high pollution. 17,18 Specifically, in order to activate the inert dinitrogen molecules, a high reaction temperature (300 C-500 C) and high pressure (200-300 atm) must be kept in the this process, thus leading to high energy consumption. According to the statistics, the annual energy consumption of the Haber-Bosch process accounts for 1%-2% of the worldwide annual energy generation. 19 Meanwhile, in order to produce the high-purity hydrogen gas required for this process, a vast amount of CO 2 is also emitted, which contributes to approximately 1.5% of the total greenhouse gas emission. 20 Obviously, seeking a green and efficient nitrogen fixation technology is necessary for the sustainable development of energy and the environment. Recently, lots of studies have indicated that photo(electro)catalysis processes can achieve effective nitrogen fixation using nitrogen gas and water as raw materials at room temperature and atmospheric pressure. Generally, photo(electro) catalytic processes for nitrogen fixation include nitrogen reduction reaction (NRR) and nitrogen oxidation reaction (NOR). For the former one, nitrogen gas is reduced to ammonia. For the latter, however, dinitrogen molecules are oxidized to nitrate. The reaction not only realizes the efficient utilization of gaseous nitrogen but also produces nitrate, which has a great impact on crops and industrial production.
In contrast to the above nitrogen fixation processes, ammonia decomposition reaction (ADR) and the denitrification reaction can convert the fixed nitrogen back into a molecular state. Specifically, ADR can oxidize ammonia to form nitrogen gas, which has been applied in two important energy-related fields. On the one hand, ADR is the key process for ammonia fuel cells (AFCs). On the other hand, ammonia can also be directly decomposed to produce hydrogen gas in ADR. Because ammonia is much easier to be liquefied than hydrogen, which makes it more facile for storage and transport than hydrogen gas, it has been regarded as an ideal storage medium of hydrogen energy. As for the denitrification reaction, nitrate or gaseous NO x can be reduced into nitrogen gas. Eutrophication caused by excessive nitrate or other gaseous nitrogen oxides emission has led to significant difficulties for environment protection in various countries. Consequently, a clean and sustainable environment is desired by the people around the world. Therefore, this process is frequently used in sewage treatment, which is of great significance for environmental protection.
The above-mentioned processes are the conversions between molecular nitrogen and fixed nitrogen. However, the nitrogen species with different valence states can also be converted to each other by the photo(electro) catalysis way. For example, NO x can also be directly reduced to generate ammonia. Producing ammonia by nitrite reduction can be an alternative to the direct ammonia synthesis by nitrogen reduction. Besides, ammonia oxidation reaction (AOR) can also transform NH 3 to nitrogen oxide at a high temperature, which is essential for the ammonia nitrogen treatment in wastewater. 21 Consequently, the nitrogen-related photo(electro)catalytic reactions, such as NRR, NOR, AOR, ADR, and others, should play a critical role in human production and life, considering its sustainability features ( Figure 1). By far, a large number of original works have been reported. For such a research hotspot, a systematic review can help us clarify the development of this field and deepen our understanding of related mechanisms. To the best of our knowledge, however, there are rarely any such reviews available on the nitrogen looping by artificial catalytic reactions. This scenario thus makes it very necessary to provide a comprehensive review to discuss in detail the catalytic reactions involving nitrogen element, especially designing the efficient photo(electro) catalysts.
Herein, we focus on the latest advances in artificial nitrogen element looping, mainly on the photo(electro) catalytic transformations among N 2 , gaseous NO x , nitrate, and NH 3 . The transformation reactions involved in the above artificial nitrogen looping process will be discussed in detail, mainly including NRR, NOR, AOR, ADR, denitrification, etc. In particular, the various photo(electro)catalysts used in the aforesaid reactions will be systematically introduced. Finally, we will point out the development direction for efficient industrial nitrogen fixation and artificial nitrogen looping processes and provide an essential guarantee for the sustainable development of energy and the environment.

| CONVERSION BETWEEN NITROGEN AND AMMONIA
The transformation between nitrogen and ammonia have been widely investigated, particularly NRR and AOR. 13,22 On the one hand, NRR at ambient conditions provides an energy-saving and low-pollution method for the synthetic ammonia industry, which benefits to fertilizer and textile industry. On the other hand, AOR and denitrification show great potentials in AFCs and hydrogen storage, respectively. In this section, we will discuss in detail the above photo(electro)catalytic processes.

| Nitrogen reduction to ammonia
As above mentioned, although the nitrogen gas in the atmosphere is abundant, it is difficult to directly use nitrogen gas due to its high inertness resulted from the high bond energy of the N≡N triple-bond. The traditional Haber-Bosch process in the industry is accompanied by high energy consumption and a large amount of carbon emission. 23 Thus, for the sustainable development of the ammonia synthesis industry, lots of efforts have been made to explore novel technologies, which can achieve NRR at ambient conditions, such as photo(electro)catalytic nitrogen reduction process.
As shown in Figure 2A, the photocatalytic NRR undergoes three processes: (a) photogenerated electron/ hole pairs are generated under the illumination by incident light; (b) photogenerated electrons and holes are separated and further migrated to the surface of photocatalysts, respectively; (c) photogenerated electrons achieved at the surface carry out NRR (N 2 + 2H 2 O + 6H + + 6e − = 2NH 3 Á H 2 O). 24,25 For the electrocatalytic NRR, however, protons are produced at the interface between the anode and the electrolyte when a voltage is applied to the electrodes. Subsequently, these protons F I G U R E 1 Looping of nitrogen, nitrogen oxides and ammonia under photo(electro)catalysis will be migrated through the electrolyte to the cathode. Then they react with dinitrogen molecules on the cathode with the help of the reductive electrons, thereby reducing nitrogen gas to ammonia ( Figure 2B).

| Photocatalytic NRR
Solar-driven NRR (ie, the photocatalytic NRR) has been regarded as a promising technology for ammonia synthesis due to the inexhaustibility of solar energy and the simple reactants (nitrogen and water), which does not produce harmful species to the environment. 26,27 However, some issues still impede the goal of efficient and highly selective photocatalytic NRR. On the one hand, the solubility of nitrogen gas in water is low, and its adsorption on the catalyst's surface is very limited. On the other hand, insufficient NRR active sites on the catalyst surface limit the photocatalytic NRR performance. Besides, the low rate of proton generation and migration, and the high recombination of photoinduced charges is also harmful to the performance of photocatalytic NRR. 28,29 Consequently, to explore efficient photocatalysts is a critical step toward high-performance photocatalytic NRR. 30,31 In recent years, various materials have been investigated in order to develop advanced photocatalysts for solar-driven NRR, including oxides, sulfides, nitrides, polyoxometalates, etc. For example, as a traditional wide bandgap oxide, TiO 2 shows a capacity of NRR due to its suitable band edges for NRR. 32,33 However, the absence of active sites and insufficient visible light absorption are still the main bottlenecks, which limit its performance enhancement. To solve these problems, oxygen vacancies (V O ) have been introduced into the TiO 2 matrix. For example, a commercially available TiO 2 with a large number of V O delivered a NH 4 + production rate of 0.76 μM/hr after the UV-light irradiation (λ = 280-420 nm, light intensity = 35 W/m 2 ) of 48 hours ( Figure 3A). 33 And the ammonia yield decreased when the number of V O reduced. Besides, Cu 2+ doping can introduce V O into TiO 2 nanosheets (NSs), thus leading to not only more NRR active sites but also enhanced visible light absorption ( Figure 3B). 34 As a result, at the optimum doping amount (6 atm%), Cu-doped TiO 2 NSs showed a high NH 3 evolution rate of 78.9 μmol/g/hr under full solar irradiation, which is almost 4.2-fold higher rate than that of pristine TiO 2 NSs.
As an emerging photocatalyst, BiOBr has become a research hotspot due to its good visible light response. [35][36][37][38] Similarly to TiO 2 , nevertheless, the deficiency of active sites is also a significant issue for BiOBr to achieve the excellent photocatalytic NRR. Fortunately, V O introduction is also a feasible strategy for BiOBr to improve its photocatalytic activity ( Figure 3C). For instance, V O defects can greatly enhance the NRR performance of BiOBr NSs from 5.75 to 54.70 μmol/g/hr. 39 Apart from oxides, nitrides and sulfides can also be utilized as photocatalysts for NRR. For example, graphitic carbon nitride (g-C 3 N 4 ) NSs etched by KOH showed an enhanced photocatalytic performance for NRR. Specifically, the KOH-treatment can bring rich defects to g-C 3 N 4 NSs, which is beneficial to promote the N 2 adsorption, light energy absorption, and charge separation, resulting in a high ammonia production rate of 3.632 mmol/g/hr ( Figure 3D). 40 Activation of N 2 is an uphill process involving the generation of high energy intermediates (N 2 H, N 2 H 2 etc.). Fortunately, the thermodynamic barrier of multielectron reactions can be reduced by avoiding the production of high-energy intermediates. As an n-type semiconductor, the ultrathin MoS 2 NSs possessed sufficient free electrons, which can be attracted by the light-induced excitons (electron-hole F I G U R E 2 Schematic illustration of (A) photocatalytic and (B) electrocatalytic NRR, respectively pairs) to form charged excitons (ie, trions, Figure 3E) under solar light irradiation. These trions have multiple electrons in one bound state, which could accelerate the multielectron transfer reactions. Consequently, the ultrathin MoS 2 NSs could reduce the thermodynamic barrier of NRR, thus leading to a relatively high ammonia synthesis rate as 325 μmol/g/hr. 41

| Electrocatalytic NRR
In addition to the photocatalytic route, the electrocatalytic process can also realize nitrogen reduction to ammonia under ambient conditions. When the electrocatalytic NRR is driven by the electric power generated from renewable energy sources (eg, solar energy, wind energy, tide/wave energy, biomass energy, etc.), this technology seems to be an attractive and environment-friendly strategy to the ammonia synthesis industry. 42,43 For the electrocatalytic NRR, there are two major issues: the low activity and the poor selectivity of electrocatalysts, which reduce the NRR performance. 23 On the one hand, the lack of active sites on the surface of electrocatalysts leads to low activity. 44 On the other hand, in the process of nitrogen reduction, the competitive hydrogen evolution reaction (HER) decreases the Faradaic efficiency (FE), leading to poor selectivity. 1,45 So as to obtain superior NRR performance, the optimization of the electrocatalysts proves distinctly necessary. 46 In order to achieve a high activity and selectivity for electrocatalytic NRR, many types of advanced electrocatalysts have been investigated, including metals (eg, Fe, 47,48 Mo, 49,50 Cu, 51 Ru, 52,53 Rh, 54 Pd, 55 Au, 56 69 VP, 70 etc), and so on.
Among the aforementioned various NRR electrocatalysts, metal elements have been widely studied, especially metallic single-atom catalysts (SACs), due to their super-high atom utilization and excellent NRR performance. 71,72 For example, Ru SACs dispersed on nitrogendoped carbon (Ru SACs/N-C) showed an ammonia yield rate of 120.9 μg NH 3 =hr=mg cat and FE of 29.6% at −0.2 V versus reversible hydrogen electrode (RHE), 52 respectively ( Figure 4A). With the hydrothermal treatment of a precursor solution containing glacial acetic acid, ZrCl 4 , H 2 BDC, and RuCl 3 , UiO-66 (Zr 6 O 4 (OH) 4 (BDC) 6 ; BDC, 1,4-benzenedicarboxylate) with confined Ru single atoms F I G U R E 3 A, The photocatalytic process for N 2 fixation on the (110)  was synthesized. After annealing, Ru@ZrO 2 /NC catalyst was obtained. The Ru sites with oxygen vacancies are claimed to be the major active centers for nitrogen reduction, and the ZrO 2 could effectively suppress the hydrogen evolution reaction (HER), resulting in an ammonia generation rate of 3.665 mg NH 3 =hr=mg cat at −0.21 V versus RHE and FE of 21% at an overpotential as low as 0.17 V versus RHE, respectively ( Figure 4B). 53 Despite of many merits, the rare resource and the high cost still lead to difficulty in the large-scale utilization of noble metal catalysts, such as Ru. Therefore, exploiting non-noble metal catalysts has been regarded as a feasible strategy to realize commercial ammonia synthesis. For instance, Han et al. prepared N-doped porous carbon (NPC) supported single Mo atoms (SA-Mo/NPC) electrocatalyst for NRR. The ammonia yield reached to 34 ± 3.6 μg NH 3 =hr=mg cat and the FE is 14.6% ± 1.6% at −0.3 V versus RHE, respectively. This superior performance originated from the dispersed single-atom active sites and 3D hierarchical structure of such catalyst ( Figure 4C). 73 Very recently, our group developed Fe/Cu atom clusters confined in sub-nanoreactors of g-C 3 N 4 on the surface of carbon nanotubes (CNTs) as efficient NRR electrocatalysts. Due to the synergistic enhancement of Fe and Cu, these Fe/Cu clusters delivered an ammonia production rate of 8.6 μg NH 3 =hr=mg cat and FE of 18.8% at −0.3 V versus RHE, respectively ( Figure 4D). 74 Apart from metal elements, metal oxides and nitrides also exhibited good performance for NRR. For example, we synthesized Fe-doped W 18 O 49 nanowires on carbon fiber papers through a facile solvothermal route ( Figure 5A). 75 This electrocatalyst exhibited a yield rate of 24.7 μg NH 3 =hr=mg cat for ammonia synthesis and FE of 20% at −0.15 V versus RHE. The DFT calculation demonstrated that effective catalytic activity is attributed to the addition of Fe, which not only increases the number of oxygen vacancies and W active sites but also optimizes the adsorption of nitrogen. Besides, Cr 2 O 3 -based electrocatalysts also achieved a high NH 3 yield of 25.3 μg NH 3 =hr=mg cat as well as a high FE (6.78%) at −0.9 V versus RHE ( Figure 5B). 61 As for metal nitrides, VN nanowires array on carbon cloth (VN/CC) exhibited an ammonia production rate of 2.48 × 10 −10 mol/s/cm 2 and a FE of 2.25% at −0.30 V versus RHE ( Figure 5C). 65 Through introducing nitrogen vacancies acting as the active sites on two-dimensional (2D) layered W 2 N 3 nanosheets (NV-W 2 N 3 ), Jin et al. 76 obtained a steady ammonia generation rate of 11.6 ± 0.98 μg NH 3 =hr=mg cat and FE of 11.67% ± 0.93% at −0.2 V versus RHE for 12 cycles of 24 hours ( Figure 5D).
As for carbon materials, doping has been considered a common method to modify the surface electronic structures of materials. [77][78][79][80] For instance, carbon nanospheres co-doped by B and N (BNC-NS) obtained a significantly improved catalytic activity of electrochemical reduction of N 2 to NH 3 . 77 The BNC-NS demonstrated the superior performance of NRR with a FE of 8.1% and ammonia yield rate of 15.7 μg NH 3 =hr=mg cat , which were attributed to the high N 2 adsorption capacity ( Figure 5E). Table 1 summarizes the NRR catalysts performance mentioned above.

| Ammonia oxidation and decomposition
With high energy density, facile storage and transportation, and carbon-free feature, ammonia is regarded as a feasible carrier for renewable energy sources. In particular, compared with the high-cost liquid hydrogen, ammonia shows facile storage and transport in the forms of the liquid or concentrated aqueous solution. In this field, ammonia oxidation reaction (AOR) and ammonia decomposition play critical roles in AFCs and ammoniabased hydrogen storage, respectively.

| Ammonia oxidation reaction for direct ammonia fuel cell
AOR is the half-reaction for direct ammonia fuel cells (DAFCs), which can generate electricity using ammonia and oxygen gas as the electrode materials. 81 Generally, a DAFC is constructed by three parts: an anode, a cathode, and an electrolyte. In its operation, ammonia gas is fed into the anode of DAFC, and reacts with hydroxide (OH − ) to produce electrons, nitrogen gas, and water (ie, AOR) (Equation (1)). At the same time, oxygen gas is fed in the cathode and then reacts with water and accepts electrons conducted from the anode to generate OH − (Equation (2)). By this means, the chemical energy stored in ammonia gas and oxygen gas is converted to electricity. Besides, the overall reaction of the DAFC system is also described in Equation (3), indicating that the environmentally friendly nitrogen gas and water are the final products.
Cathode : Overall : Although DAFCs exhibit excellent potential, the slow kinetics process of AOR hinders the further improvement of DAFCs. For solving this problem, a serial of highperformance catalysts has been designed to accelerate the AOR, mainly including metal elements, binary, and ternary alloy catalysts. [82][83][84][85][86] For example, with a polycrystalline gold disc electrode (5 mm diameter) acting as the current collector, the preferentially oriented Pt (100) nanoparticles (NPs) were fabricated through a simple colloidal route, exhibiting a higher current density than polycrystalline Pt ( Figure 6A). 87 The superior performance of the preferentially oriented Pt NPs can be ascribed to the high sensitivity of AOR to surface sites with square symmetry.
Except for monometal elements, the bimetal catalysts and ternary alloy catalysts are also active for AOR. For example, Pt/Ru (atomic ratio of 90:10) alloy NPs exhibited lower onset potential than Pt ( Figure 6B). 88 In another  Figure 6C). 89 As for the ternary alloys, Pt-Ir-Ni NPs dispersed on the porous silicon dioxide (SiO 2 ) and carboxyl-functionalized carbon nanotube (SiO 2 -CNT-COOH) exhibited high catalytic activity for the AOR, which was evidenced by its low onset potential ($0.40 V vs RHE) at room temperature, which can be attributable to abundant OH ad generated by porous SiO 2 and the augmented conductivity by CNTs ( Figure 6D). 90 Table 2 Summarizes the AOR catalysts performance mentioned above.

| Ammonia-based electrocatalytic hydrogen storage
As an ideal carbon-free fuel, hydrogen gas has been widely studied as the new energy carrier due to its high energy density of 146 MJ/kg. [91][92][93][94] However, the largescale utilization of hydrogen is still limited because of the high costs resulting from the transportation and longterm storage of hydrogen. Specifically, the boiling point of hydrogen is as low as −253 C under standard atmospheric pressure and its Carnot efficiency is 7.3%, which means that the liquefaction of hydrogen requires high energy. 95 In addition, the hydrogen is liable to boil off, which leads to the gradual loss of hydrogen with the increase of storage time. Fortunately, ammonia is easier to store and transport and can decompose to hydrogen without carbon emission, which makes it a feasible medium of hydrogen storage. In addition, electrocatalytic ammonia decomposition to generate hydrogen gas can be achieved under low-temperature conditions (<450 C), which facilitates the large-scale implementation of this technology. 14,16,96 As shown in Figure 7A, the process from storing ammonia to release energy from hydrogen contains five steps: ammonia storage, ammonia decomposition, hydrogen separation and purification, hydrogen storage and compression, as well as last hydrogen use. 16 Among them, the ammonia decomposition process will be mainly discussed in this section. Ammonia decomposition occurs in a series of steps of dehydrogenation reaction(Equations (4)- (7)). The reaction starts from the adsorption of ammonia in the beginning and decomposition to N 2 and H 2 at last (Equation (8)). The same as NRR, nitrogen adsorption is also regarded as the rate-limiting step for ammonia decomposition.
where, X * represents the adsorbed species.
In recent years, many materials have been utilized as catalysts for ammonia decomposition, such as metal catalysts (including metal 97,98 and alloys 99,100 ) and alkali metal amides. 101 For example, The graphene and defective graphene-supported standing triangular Pt 3 exhibited high reactivity and low activation barriers, which was attributed to the existence of Lewis acid/base pair sites accommodate the adsorption and subsequent dissociation of NH x * ( Figure 7B). 102 Apart from Pt, Ru also exhibits a good catalytic activity for ammonia decomposition. Specifically, the mesoporous Ru/MgO composite was fabricated by a deposition-precipitation method. Due to its high surface area and porous structure, this catalyst showed a high electrocatalytic activity of ammonia decomposition. The Ru/MgO catalyst presented the H 2 evolution rate of 21.2 mmol/min/g, which is higher than that of 14.5 mmol/min/g on Ru/CNTs, meaning that its own an excellent performance of electrocatalytic ammonia decomposition. Further, the catalytic activity of the sample was enhanced to 35 mmol/min/g after a KOH treatment due to the efficient electrons donation of K + ions ( Figure 7C). 103 In addition, the alloy catalysts are also attractive. For example, Ni-Co bimetallic catalysts were used to decompose ammonia and showed the turnover frequency (TOF H 2 ) as high as 40.57 min −1 at 350 C ( Figure 7D). 104 The Cs-modified Co 3 Mo 3 N catalysts were synthesized by a facile single-step method, which improves the catalytic activity for ammonia decomposition and promotes the recombinative desorption of hydrogen and nitrogen atoms ( Figure 7E). 105 Apart from metal catalysts, catalyzing the ammonia decomposing process via alkali metal amides is another feasible method. For instance, accelerated decomposition of ammonia was achieved through the synchronous stoichiometric decomposition and sodium amide (NaNH 2 ) regeneration. As a result, the decomposition efficiency of 90% was obtained at 500 C using 0.5 g NaNH 2 in a 60 sccm NH 3 flow, which is higher than that of the Ru (82%) and Ni (58%) catalysts. 106 Table 3 summarizes the ADR catalysts performance mentioned above.

| CONVERSION BETWEEN NITROGEN AND NITROGEN OXIDES
The conversion between N 2 and NO x is the key step during the whole process of nitrogen cycling. On the one hand, NOR is another effective way to fix nitrogen, which realizes the conversion from nitrogen to nitrate. 107 On the other hand, gaseous nitrogen oxides (eg, NO and NO 2 ) are the primary pollutants in the air, which should be removed or converted to nitrogen. Meanwhile, the eutrophication and pollution of water caused by excessive nitrate also should be paid attention to. [108][109][110] For example, waste gas and sewage treatment are very important and attracting widespread attention. 111 Consequently, the NOR and reduction of nitrogen oxides are still hot research topics to be further studied.

| Nitrogen oxidation reaction
As mentioned above, NRR is an effective way for nitrogen fixation. In addition to this route, nitrogen oxidation to nitrate (ie, NOR) has also been considered a feasible way to achieve the same goal of fixing nitrogen. Different from NRR, however, the products of NOR are nitrates rather than ammonia, which can not only provide nitrogen sources and nutrition for soil but also play a crucial role in numerous manufactures, for example, making gunpowder, curing coronary heart disease, and so on. 108,[112][113][114][115] It has been suggested that N 2 is oxidized to nitrate by the following two steps: the first step is to activate the N≡N triple-bond, converting the inert N 2 to active NO * intermediates ( * represents to active sites). The second step is a redox reaction, in which NO * is transformed into nitrate over the active sites. 116 In this process, the highly active and selective catalyst materials for NOR are very important.
In order to develop efficient catalysts for NOR, significant efforts have been made. The N≡N triple-bond activation for the direct synthesis of nitrate is very important because no other side effects or by-products, which reduce the energy efficiency, will occur in this process. 117 In the photocatalytic NOR process, the valence state of nitrogen is increased due to the oxidation effect of the light-induced hole. Consequently, some typical photocatalysts, such as TiO 2 , have been adopted for this purpose. 118 However, the activation of nitrogen molecules is extremely difficult. 119,120 To tackle with this issue, some researchers have utilized the electron-rich regions on the surface of the catalyst, such as the surface pits or vacancies, to directly absorb nitrogen molecules and promote the charge transfer. 121,122 Therefore, catalyst materials for nitrogen oxidation are generally designed to improve their catalytic performance through structural design, such as the design of atomic configurations. Catalysts with local electron-rich regions and uneven surface structures would produce a synergistic catalytic effect for the adsorption and dissociation of nitrogen molecules, which is expected to overcome the problem of difficult nitrogen activation.  Figure 8A). 117 ZnFe x Co 2−x O 4 spinel oxides were also utilized as catalysts to carry out electrocatalytic NOR, 123 which achieved the direct synthesis of nitrate with high yield (130 ± 12 μmol/ hr/g MO at 1.6 V vs RHE) by iron catalysts (Figure 8B,C). Besides, modified MXene electrocatalysts have also been utilized as electrocatalyst for NOR, which achieved 2.80 μg/hr/mg cat yield of nitrate ( Figure 8D). 124 In the meantime, Ru-doped TiO 2 /RuO 2 has shown a decent catalytic activity for NOR as well, and its FE is 26.1%. Density functional theory calculation showed that Ru δ+ could be the main active center in NOR ( Figure 9A). 116 Besides, the Fe-doped SnO 2 can also achieve an efficient NOR ability. Specifically, the NO 3 − yields and FE of modified SnO 2 at the Fe-doping amount of 3% achieved 42.9 μg/hr/mg cat and 0.84%, respectively. Further, there are two existing forms of Fe element in SnO 2 matrix, Fe single atoms anchored on Vo and lattice doped Fe, which improve the absorption and activation of N 2 and promote the electrical conductivity of SnO 2 ( Figure 9B). 125

| Denitrification of gaseous nitrogen oxides
Since the first industrial revolution, large-scale human activities, such as fossil fuel combustion, automobile exhaust, industrial nitrate production, and artificial nitrogen fertilizers, have led to a serious imbalance in the global nitrogen cycle. 126,127 The excessive emission of nitrogen oxides, such as NO and NO 2 , is not only harmful to the natural environment but also poses a threat to the sustainable development of human society. Therefore, the treatment and control of such nitrogen oxides in the air have become an urgent issue that remained unsolved. In order to solve these issues and alleviate environmental and health problems caused by air pollution, denitrification treatment based on photo(electro)catalysis to reduce pollutants in waste gas through catalysis has attracted extensive attention. Photocatalytic denitrification treatment shows great potential in the reduction of nitrogen oxides in the exhaust gas. In particular, Ti-based materials have been extensively studied due to their high photocatalytic activity and stability. 116,[128][129][130][131][132] Titanium alkoxides prepared by the high-temperature hydrolysis method have shown high denitrification activity, with a high absorption rate of 98% for NO photocatalytic oxidation. 133 In 2008, TiO 2 nanoparticles coating on woven glass fabric was used as a photocatalyst, which showed a good effect for removing nitrogen oxides and an efficiency of 63.9% upon the steady-state. Two technologies, including photocatalytic oxidation and wet absorption, were jointly utilized to deal with nitrogen oxides in the exhaust gas, which greatly improved NO x removal rate up to 75%. They used photocatalytic oxidation convert NO to NO 2 at the first step, then applied a sulfite solution as an absorbent to efficiently absorb nitrogen oxides ( Figure 10A); and the main products obtained through this process are nitrite and nitrate, which can effectively reduce air pollution gas. 134 From 2013 to 2014, based on the selection of photocatalytic materials, researchers conducted relevant model construction and theoretical calculation, controlled and described the photocatalytic oxidation reaction of NO, and deduced the rate law of NO oxidation, providing data support for the control of NO oxidation rate. 128,135 Recently, the composite material based on the novel microporous MIL-100(Fe)/Ti 3 C 2 MXene structure could not only increase the material's thermal stability under visible light but also boost the activity of nitrogen fixation ( Figure 10B). The synergistic effect of the Schottky junction makes the photocatalytic activity greatly enhanced, which is expected to treat NO and reduce air pollution efficiently. 136 From the above discussion, it can be seen that gaseous nitrogen oxides (NO x ) will be converted into nitrate during the photocatalytic process. Further, with the deepening of research, the performance of the catalysts has been improved to a large extent, including the design of F I G U R E 9 A, The NOR mechanism of Ru/TiO 2 composite electrocatalysts. Reproduced with permission: Copyright 2020, Wiley-VCH. 116 B, Fe-doped SnO 2 structure for NOR. Reproduced with permission: Copyright 2020, Wiley-VCH 125 catalysts structure and the improvement of the activity. It is a feasible method with a very promising prospect for treating large quantities of waste gas. However, there are still very limited studies on the conversion of nitrogen oxides to N 2 or nitrate by electrocatalysis. In the future, it might be a very important topic to explore the electrocatalysis route for waste gas treatment.

| Nitrate reduction
In the above discussion, the transformation from nitrogen oxides to nontoxic nitrogen gas is necessary. But in fact, more nitrogen oxides are oxidized to produce nitrate so as to reduce pollutants in the air. However, the excessive soluble nitrate species in the water causes severe problems for the water resource. Therefore, it is necessary to reduce the excess nitrates. If the nitrates can be effectively reduced, it will bring a serial of benefits for the environmental sustainability. Effective approaches to reducing nitrate to N 2 by photo(electro)catalysis to achieve green and the carbon-free conversion will be discussed in detail.

| Electrocatalytic nitrate reduction
Electrocatalytic nitrate reduction reaction can be regarded as the most promising nitrate conversion route. 136 Specifically, noble metal electrocatalysts, especially alloys, exhibited excellent performance for nitrate reduction. [137][138][139] For instance, bi-metallic alloys of Pt/Rh and Pt/Ir have been utilized as catalysts in the electrocatalytic nitrate reduction reaction. Remarkably, the nitrate reduction performance of alloy catalysts (ie, Pt/Rh and Pt/Ir) was superior to the metal-element catalysts of Pt, Rh, and Ir. 140 Those above-mentioned expensive materials, however, impose restrictions on the development of electrocatalytic nitrate reduction, which limited its industrial deployment. Consequently, those non-noble counterparts have been explored to carry out the efficient and stable nitrate reduction process. 141,142 For example, a corchorifolius-structure composite material ( Figure 11A), in which Fe nanoparticles were encapsulated in a carbon microsphere with a rough surface, has been developed. The novel materials was used as electrocatalysts for nitrate reduction. Importantly, a high removal ability (1816 mg/N g) and an excellent nitrogen selectivity (98%) were successfully achieved. 143 In another research, FeN NPs coated by thin-layer carbonized nitrogen (NC) were synthesized ( Figure 11B). Importantly, this novel electrocatalyst of NC@FeN possessed a high selectivity (91%) and nitrate removal rate (6004 mg/N g Fe). 144 Apart from the iron-containing catalysts, Cu-based composites also demonstrated unique advantages for the nitrate reduction reaction. 145,146 For example, a nanosized composite that consists of monodisperse Cu nanoparticles and reduced graphene oxide (rGO) has been used as the catalyst for electroreduction of nitrate ( Figure 11C). This composite catalyst delivered a nitrate removal rate of approximately 96.8%, which was several times higher than GP (~23.27%) and Cu/GP (~82.79%) electrodes. 147

| Photocatalytic nitrate reduction
Photocatalytic technology can effectively remove the toxic constituents in the environment, such as organic dyes and bacterias, which has been regarded as an ecofriendly strategy for nitrate reduction. 148,149 Recently, breakthroughs have been made in the research of nitrate reduction via the photocatalytic route. For example, a novel Pd/GdCrO 3 composite material was prepared and used as the photocatalyst to carry out photocatalytic nitrate reduction reaction ( Figure 12A). The Pt content in Pd/GdCrO 3 played an important role on photocatalytic performance. The composite catalyst with 1 wt% of Pt amount owned the highest nitrate removal rate (98.7%) and selectivity for N 2 (100%). This was caused by not only the negative conduction band value of GdCrO 3 but also the catalysis effect of Pd. 150 Besides, AgCl-modified TiO 2 nanotubes (AgCl/TNTs) also owned good photocatalytic activity of nitrate reduction. Specifically, AgCl/TNTs showed an excellent reduction effect of nitrate (nitrate reduction 94.5% and N 2 selectivity 92.9%) for 30 minutes under 365 nm UV irradiation with the formic acid as hole scavenging agent ( Figure 12B). 129 F I G U R E 1 1 A, Like Corchorifolius structure iron-carbon microspheres. Reproduced with permission: Copyright 2019, American Chemical Society. 143  The following table mainly combs the performance of catalysts used in nitrogen oxidation reaction and nitrate reduction to nitrogen (Table 4).

| CONVERSION BETWEEN AMMONIA AND NITRIC ACID/ NITRATE
The conversion between ammonia and nitric acid/gaseous nitrogen oxides mainly involves two redox routes: the ammonia oxidation to nitrogen oxides and the direct reduction of nitrate to ammonia. For the former one, it is a key step in industrial nitric acid production; while, for the latter, it opens a new way to synthesize ammonia route from soluble nitrate instead of inert nitrogen. In this section, we will discuss these processes in detail, with special emphasis on the advanced catalysts used in these two artificial photo(electro)catalystic processes.

| Ammonia oxidation in the Ostwald process
At present, the production of nitric acid mainly depended on the traditional Ostwald process. This way includes three critical steps: (a) ammonia is oxidized to NO with the help of Pt/Rh catalysts; (b) NO is further oxidized to NO 2 ; and (c) NO 2 is absorbed by H 2 O to produce nitric acid. 151 In these successive steps, ammonia oxidation (ie, steps a and b) is the most critical. Therefore, the ammonia oxidation in the Ostwald route should be paid more attention to.
As mentioned, precious metals, such as Pt and Rh, are frequently used catalysts for ammonia oxidation during the Ostwald process. Therefore, searching for efficient and low-price catalysts to replace these noble metal-based catalysts is necessary for the nitric acid synthesis industry.
In nature, the nitrification of ammonia to NO 2 − takes place through the followed steps: (a) the ammonia monooxygenases catalyze ammonia oxidation to hydroxylamine (NH 2 OH); (b) the oxidoreductases promote the fourelectron oxidation of NH 2 OH to produce NO 2 − . 152-154 Inspired by microbial route in nature, a biomimetic method of ammonia oxidation to NO 3 − can be achieved under the effect of enzymes exiting in ammonia-oxidizing bacteria. 155 Specifically, a microbial biofilm was incubated at 46 C in an ammonium-containing mineral medium to enrich moderately thermophilic AOM. In the mineral media, ammonium acts as the sole source of energy and reductant, and bicarbonate/CO 2 acts as the sole carbon source. And then, the data of near-stoichiometric oxidation ammonium to nitrates is obtained in the enrichment culture ENR4. The nitrospria bacteria can achieve the nearly complete oxidation of the 1 mM ammonium to nitrate ( Figure 13A).

| Direct reduction of nitrate to ammonia
As above-mentioned, the traditional Haber-Bosch process requires a huge energy input as well as leads to significant carbon emissions. To solve this issue, NRR at the ambient condition has been widely studied. However, the low solubility in water and high inertia of nitrogen gas will lead to two serious problems: low yield and poor selectivity. The direct reduction of nitrate to ammonia could provide a more promising strategy for ammonia industry in the future due to higher solubility and activity of nitrate in aqueous solution than that of nitrogen gas. 156 165 etc.), and so on. Early research has revealed that Al showed a high selectivity of 12% in the process of direct reduction of nitrate to ammonia ( Figure 13B). 159 Apart from metal elements, metal oxides also show great potential in the direct electroreduction of nitrate to ammonia. Specifically, Co 3 O 4 /Ti was utilized as electrocatalysts to carry out the above reaction. And it is observed that the NO 3 − -N removal and NH 4 + -N generation efficiencies are dependent on the initial NO 3 − -N concentration. When the initial concentration is 50 mg/L, the NO 3 − -N removal efficiency is above 85%; however, becoming lower than 20% when the initial concentration increases to 500 mg/L. ( Figure 13C). 163 Very recently, a Cu-containing crystalline (3,4,9,10perylenetetracarboxylic dianhydride [PTCDA]) was synthesized and utilized as the electrocatalyst for direct nitrate reduction to produce ammonia. The FE of PTCDA reached a maximum value of 85.9%, and the rate of NH 3 generation was 436 ± 85 μg/hr/cm 2 at −0.4 V versus RHE. 160 The outstanding performance of this catalyst can be attributed to two aspects. On the one hand, suppressed HER and enhanced H-N bonding, which originates from Cu electronic structure. On the other hand, the microstructure of PTCDA is favorable for the transport of electrons and protons to Cu active sites ( Figure 13D). The following table mainly summarizes the performance of electrocatalysts used in nitrate reduction (Table 5).

| SUMMARY AND PERSPECTIVE
The photo(electro)catalytic reactions during artificial nitrogen looping play a crucial role in many aspects of modern society, including the chemical industry, environmental protection, clean energy source, and so on. Advanced catalysts involved in the above processes are essential for realizing efficient, stable, and low-cost artificial nitrogen looping. In this review, we have summarized the latest progress in this hot spot, and further provided an in-depth discussion on the rational design of these state-of-the-art photo(electro)catalysts.
As abovementioned, the development of photo(electro) catalysts used in the artificial nitrogen looping can bring a great breakthrough in the fields of energy and the environment. However, there are still a series of key challenges remaining unsolved, such as low efficiency, high cost, poor stability, and absence of in-depth theoretical understanding. Thus, more efforts should be made on some key points in order to further promote the development of these catalysts, and finally achieve the practical application, as following ( Figure 14): 1. High-performance photo(electro)catalysts. The high efficiency and good selectivity are two important aspects of a high-performance photo(electro)catalytic reaction. In the field of artificial nitrogen looping, however, the current researches focus more attention on designing and conceptual demonstration of novel catalysts. The efficiency and selectivity of these catalysts are still relatively low, especially for the emerging NRR. 21 In order to achieve the goal of practical application, the efficiency and the selectivity of these photo(electro)catalysts should be further improved by means of some feasible strategies, such as exposing more active sites, regulating the electronic structure of catalysts, designing electrode materials with highly ordered nanoarray architectures, and so on. 2. Low-cost and stable photo(electro)catalysts. Low cost and good stability are also crucial points for the commercial application of photo(electro)catalysts in artificial nitrogen looping. At present, noble metals are still mainly used as catalysts in artificial nitrogen looping. However, the rarity and the high value of noble metals will limit the large-scale implementation of artificial nitrogen looping by means of photo(electro) catalytic routes. Consequently, searching the earthrich and cheap alternatives, such as transition metal elements/alloys/oxides, carbon-based materials, etc, is an urgent task in the future. [137][138][139] Meanwhile, enhancement on the durability of catalysts in full lifecycle is equally important as an improvement in their performance. 166 3. In-depth theory research. Assisted by some theoretical calculations (eg, density functional theory), the catalytic mechanisms can be well understood at the atomic scale. For instance, kinetic study (eg, Langmuir-Hinshelwoodtype rate law) can reveal the process and the intermediates of photocatalytic oxidation of NO, which can help us to optimize the reaction. 128,135 Besides, kinetic and thermodynamic processes of photo(electro)catalysis can also be clearly manifested by some theoretical study. All of these are very helpful for us to deeply understand the mechanisms of artificial nitrogen looping and design the high-performance photo(electro)catalysts.
To sum up, the photo(electro)catalytic artificial nitrogen looping and corresponding catalysts have attracted increasing attention due to great potentials in the future industry, agriculture, environmental protection, and sustainable energy source. Although there are still many obstacles on the road toward the goal of commercialization, we believe that artificial nitrogen looping will play a key role in the future and will help pave the way for the sustainable development of society.