Progress of Metal Oxide (Sulfide)-Based Photocatalytic Materials for Reducing Nitrogen to Ammonia

*e Haber–Bosch process has been an important approach to produce ammonia for meeting the food need of increasing population and the worldwide need of nitrogenous fertilizers since 1913. However, the traditional ammonia production process is a high energy-consumption process, which usually produces 1 metric ton ammonia with releasing around 1.9 metric tons CO2. Photocatalytic ammonia synthesis under solar light as energy source, an attractive and promising alternative approach, is a very challenging target of reducing fossil energy consumption and environmental pollution. *erefore, photocatalytic ammonia production process would emerge huge opportunities by directly providing nitrogenous fertilizers in a distributed manner as needed in the agricultural fields. In this article, different metal oxide (sulfide)-based photocatalytic materials for reducing nitrogen to ammonia under ambient conditions are reviewed. *is review provides insights into the most recent advancements in understanding the photocatalyst materials which are of fundamental significance to photocatalytic nitrogen reduction, including the state-of-the-art, challenges, and prospects in this research field.


Introduction
As the important chemicals to our planet, nitrogen(N 2 ) compounds, such as ammonia (NH 3 ), nitrates, and urea, have played an essential role in meeting the growing demand for food and the worldwide need of nitrogenous fertilizer since 1913 [1]. It is essential for all living organisms to being provided the N 2 compounds needed for growing tissue [2]. Moreover, the global N 2 cycle, whose small fraction is related to the atmospheric ionization, mainly depends on the biogeochemical cycles [3]. In addition, N 2 photoreduction to nitrogenous compounds (e.g., NH 3 ) in soils and sands as catalysts also is an important part of global N 2 cycle. rough continuous exploration of nitrogen fixation, the Haber-Bosch process, a thermo-chemical catalytic conversion technology, becomes a primary choice of synthesizing N 2 fixation compounds which were produced from the reaction: In the past over decades [4], the process has fed the world's growing population and has significantly contributed to the human beings development. e Haber-Bosch process reacted with the pure feed gases at high temperatures and pressures, requiring almost 2% of global energy consumption [5]. On the contrary, there are some negative impacts of the Haber-Bosch process in terms of the overuse of nitrogenous fertilizer that affects the ecosystem balance, human health, climate change, etc. [6]. en, reducing the fossil energy consumption and environmental pollution would be a very challenging target for this process.
Ambient reduction of N 2 to NH 3 has been a significant research subject of efficiently fixing N 2 under mild conditions. Energy coming from sustainable source as solar, an alternative, sustainable NH 3 synthesis process based on the biological N 2 fixation would be more energy efficient than the Haber-Bosch process [7]. Additionally, the distribution of nitrogenous fertilizers produced by the Haber-Bosch process requires efficient transportation which may be least easy in the nonindustrialized countries than in the industrialized countries. en, it would be possible to produce nitrogenous fertilizers close to the farm field as needed and reduce greenhouse gas emission and control the global N 2 cycle. Obviously, due to its energy saving and friendly environmental, photocatalytic reduction of N 2 to NH 3 could be an excellent alternative to the Haber-Bosch process. at is to say, developing efficient photocatalysts for synthesizing NH 3 would emerge huge opportunity to directly provide nitrogenous fertilizers in agricultural fields as needed in a distributed manner. Chemists should discover how to activate the N≡N bond (941 kJ·mol −1 ) to synthesize ammonia in the presence of novel photocatalysts with less fossil energy consumption and more solar energy [8].
In initial works, there was a popular perspective that the photocatalytic reduction of N 2 to NH 3 should take place over abundant soil minerals and sand in nature [9]. Since 1977, the TiO 2 -based synthetic materials as photocatalysts were firstly applied for N 2 fixation under UV light, and a series of studies [10] have gradually been conducted on photocatalytic N 2 reduction to NH 3 with water and air under light driving force. Although there have been many new photocatalysts and approaches to solve the problems of the Haber-Bosch process, till now, no viable and efficient catalysts could match all requirements of an active, selective, scalable, long-lived catalyst for the sustainable photocatalytic reduction of N 2 to NH 3 [11]. is review tries to make them possible to ascertain a comprehensive understanding of the current knowledge regarding to photocatalytic materials for producing nitrogenous fertilizers from N 2 and to give a solid research perspectives for the next step work in the research field.

Photocatalytic Reduction Process of N 2 to NH 3
e photocatalytic reduction of N 2 to NH 3 under ambient conditions is a sustainable NH 3 production process without fossil energy consumption and environmental pollutions. Photocatalysis on semiconductors is a promising method for synthesizing NH 3 from N 2 with water (H 2 O) as a reducing reagent. e photocatalytic NH 3 production process mainly consists of the photocatalytic oxidation of H 2 O to protons and the photocatalytic reduction of N 2 to NH 3 .
e reactions mentioned above imply that the photocatalysts created on the surface of robust semiconductors capable of oxidizing H 2 O (Equation (2)) may reduce N 2 by the photoformed conduction band (CB) electrons (Equation (3)). As a result of these behaviours, N 2 is reduced to NH 3 by the CB electrons (Equation (2)). at is to say, NH 3 could be produced from N 2 and H 2 O (Equation (4)) on a metal-free photocatalyst under ambient conditions in these processes. e valence band (VB) should be lower than the oxygen evolution potential, while the position of the CB should be higher than the reduction potential of N 2 hydrogenation. With synthesizing ammonia, the metal oxide-based material as photocatalyst is dynamically converted between its oxidized and reduced states in the process. Solar light as energy source is a good choice for photocatalytic reduction of N 2 to NH 3 under ambient conditions because it is difficult for the existing power plants to meet the large energy need of producing NH 3 (Figure 1). e CB position of the semiconductor should be more negative than the reduction potential of the N 2 hydrogenation, as well as the VB should be more positive than the oxygen evolution potential (Table 1). erefore, something must be taken into account that both reduction potential of the adsorbate and position of the energy band are important for the photocatalytic redox reaction occurrence when making a decision on the choice of semiconductor photocatalyst materials.
Normally, the photocatalytic reduction process of N 2 to NH 3 can be divided into several reaction steps [13,14]. Firstly, the photogenerated electrons (e − ) are promoted to leave from vacant holes in the VB to the CB. Secondly, some electrons and holes recombine each other in order to improve the solar conversion efficiency and apparent quantum yield of the reduction. Protons could be generated in the endothermic process of H 2 O oxidation when solar energy is enough to be absorbed and converted into the chemical energy which can meet the large free energy gain of N 2 to NH 3 (⊿G 0 � 339 kJ·mol −1 ) [15]. irdly, the others transfer onto the surface of the photocatalysts for the redox reaction (Equation (4)) which the photoformed VB holes (h + ) oxidize H 2 O to protons (H + ) and O 2 under visible light. In the process of photocatalytic ammonia synthesis, the photogenerated holes must also be consumed to satisfy the charge neutrality (CB e − + VB h + � 0) (Equation (2) in Table 1).
By synthesizing NH 3 from atmospheric N 2 and H 2 O in the presence of photocatalysts under visible light (Equation (3)) [16], photons as the driving force actually can promote to reduce N 2 to NH 3 after a series of multistep injections of photogenerated electrons and H 2 O-derived protons. erefore, the first electron transfer (−4.16 V vs. NHE) and proton-coupled electron transfer (−3.2 V vs. NHE) ( Table 1) must be overcome to reduce N 2 to NH 3 . en, the recombination of charge carriers and a small band gap are important for satisfying semiconductor photocatalysts preferably in the visible light. And it is also necessary for ideal semiconductor materials to have the characteristics of the charge carrier recombination and a small band gap in the visible light region.

Photocatalytic Materials for Reducing N 2 to NH 3
As we all know, photocatalytic NH 3 synthesis from N 2 also needs to overcome the N≡N band energy barrier with the high-energy UV light. Necessarily, the large thermodynamic driving force is required for the reduction of N 2 to NH 3 at atmospheric conditions [17]. To solve the problem, developing the e cient photocatalytic materials for the N 2 reduction process maybe are indispensable parts for increasing NH 3 yields signi cantly. It is rst time that Dhar and his coworkers [18] discovered this photocatalytic reaction in the 1940s. Schrauzer and Guth [13] demonstrated that the photocatalytic NH 3 synthesis on a variety of sterile desert sands until late 1970s. Following this work, some independent results were achieved as gaining synthetic materials for speeding the photocatalytic N 2 reduction rate [19]. However, there are still some di erences among these results some of which are in con ict with each other or confused, and no consensus was reached in regard to the photocatalytic materials challenges for the reaction of N 2 and H 2 O to NH 3 and oxygen. In this section, the metal oxide-based photocatalytic materials for the reduction of N 2 to NH 3 are classi ed (Tables 2 and 3) according to their di erent elemental compositions and the state-of-the-art research advancements are described, respectively.

Titanium Oxide-Based Materials. Photocatalytic N 2
xation has received more and more attentions since synthesizing NH 3 from N 2 was available in the presence of titanium-based catalysts. Dhar and Chowdhry [43] found that titanium oxide (TiO 2 ) as photocatalytic material could provide the larger NH 3 yields from reducing N 2 than ferric oxide and zinc oxide as photocatalysts. Schrauzer and Guth [13] also discovered this phenomenon of N 2 photoreduction. UV irradiation of TiO 2 with a large number of surface Ti 3+ species could produce NH 3 from N 2 and H 2 O under ambient conditions, which shows the species are actually the active sites for the reduction of N 2 to NH 3 [14].

Iron-Doped Materials.
As key factors underlying photocatalytic performance, the in uence of the crystalline phase and iron dopants on the catalytic activity of TiO 2 photocatalytic material for the N 2 reduction was focused. Under UV light, NH 3 yields reached a maximum value when over 0.2% Fe 2 O 3 was doped into TiO 2 material. However, Hirakawa et al. [14] demonstrated that Ru, Pt, or Pd particles loaded in the TiO 2 materials were unhelpful to increase the NH 3 yields due to their particles covering the catalytic active site Ti 3+ of TiO 2 photocatalytic material and changing surface defects. Interestingly, Augugliaro et al. [44] supported Fe 2 O 3 -hybridized TiO 2 on Al 2 O 3 as photocatalyst which was applied in gas-solid uidized bed reactors to improve NH 3 yields because of the in uence of iron ions on the TiO 2 crystalline [45]. Lashgari and Zeinalkhani [21] photosynthesized NH 3 in a H 2 O photosplitting setup in the presence of some synthetic Fe 2 O 3 and TiO 2 -based uniform nanoparticles and obtained the maximum NH 3 yield. Bourgeois et al. [26] suggested that thermal pretreatments could generate surface defects or impurity states on the surface of unmodi ed TiO 2 that showed the photocatalytic activity of reducing N 2 after annealing in air. As a highly e cient method, high pretreatment temperatures (∼1000°C) inducing surface defects in TiO 2 material and the small amounts (<1%) of iron impurities are necessary to enhance the NH 3 yields [46].
Observing no reduction of N 2 to NH 3 in the presence of pure TiO 2 material, Augugliaro et al. [47] put forward a hypothesis that Fe 3+ of photoassisted organ iron compound could temporarily trap photons and promote the separation of charge carriers, which has played an important role in the NH 3 production processes. Exposing the facets of Fe-doped TiO 2 with ethanol as the scavenger enhanced the photocatalytic reduction of N 2 to NH 3 [22]. No nitrite was formed in the ethanol solution that could prevent the oxidation of NH 3 . Keeping a low doping concentration on the surface of TiO 2 material was very important to inhibit charge recombination for forming Fe 2+ and Fe 4+ and to transfer electrons and holes to Ti 4+ and OH − for generating Ti 3+ and OH [18]. Similarly, Ranjit and Viswanathan [48]    Journal of Chemistry enhance the separation of charge carriers and also demonstrated this hypothesis. However, Soria et al. [25] suggested that excess Fe dopants might lead TiO 2 materials to lose themselves all photocatalytic activities. erefore, the low bulk Fe 3+ concentrations significantly assist in charge separation. TiO 2 nanoparticles codoped with N 2 and Fe 3+ were prepared using the homogeneous precipitation-hydrothermal method. e cooperation of N 2 and Fe 3+ leads to narrow the band gap, improve the photocatalytic activity in the visible light region, and promote the separation of the photogenerated electrons and holes to accelerate the transmission of the photocurrent carrier [49]. Generally, titanium oxide-based photocatalytic materials for the N 2 reduction have made much progress in several aspects, such as the importance of iron dopants. However, some atomic-scale behaviours of TiO 2 photocatalytic N 2 reduction to NH 3 are chaos, such as the N≡N dissociation transition-states and active sites of TiO 2 photocatalytic materials. Future research works should mainly discover the secrets on how to reduce N 2 to NH 3 in the presence of titanium-based materials from a molecular-level viewpoint.

Transition Metal-Doped Materials.
Transition metals as dopants in photocatalytic materials have some advantages for the higher NH 3 yields, such as less charge carrier recombination, stronger photocatalyst absorbance, and lower overpotential for the N 2 reduction. In the process of preparing titanium-based photocatalytic materials for NH 3 synthesis, preprocessing titanium-based materials with high temperatures (≈1000°C) and adding transition metal impurities (<1%) were necessary and benefit to obtain efficient titanium dioxide-based photocatalytic materials [46]. Since then, CuCl, WO 3 , and FeOx as photocatalysts for NH 3 synthesis from N 2 have been investigated [34,50]. ose surface oxygen defects of titanium metal could play an important role in the atmospheric N 2 photoreduction.
Some transition metals taking place of iron as dopants in TiO 2 materials were investigated [13], such as Co, Mo, Ni, Pd, V, Cr, Cu, Pt, Ag, Au, and Pb. Only did Co, Mo, and Ni dopants among these metals enhance the NH 3 yields, and the other metals not. Differently, Palmisano found that chromium-doped TiO 2 material was effective in the photocatalytic N 2 reduction to NH 3 [51]. Besides their dopant   [42]. As cocatalysts of TiO 2 materials, the photocatalytic activities of transition metals are associated with the strength of the M-H bonds (M: transition metal), such as Ru > Rh > Pd > Pt [19]. Due to lower Fermi level than the others, transitionmetal dopants as electron sinks in titanium photocatalysts could minimize the probability of carrier recombination to promote greater NH 3 yields, which is a critical factor for designing photocatalytic systems. To increase N 2 reduction driving force to NH 3 synthesis, Co, Mo, Ni, Ru, and Pt dopants have been added into the titanium-based materials, respectively, which resulted in increasing the NH 3 yields [9,19]. Anatase TiO 2 was prepared by the sol-gel method through the hydrolysis of titanium tetrachloride and doped with transition metal ions like V 5+ and Zn 2+ . Although both doped samples showed similar red shift in the band gap, Zn 2+ (0.06 at.%) doped TiO 2 materials showed enhanced activity which was attributed to smaller crystallite size and larger surface area for accelerating the interfacial charge transfer [52]. Under ambient temperature and atmosphere, the photocatalytic activity of the Ag-TiO 2 multiphase nanocrystal composite thin films prepared by the liquid phase deposition method exceeded that of pure TiO 2 thin films when the AgNO 3 concentration was kept in the range of 0.03-0.05 [53]. e presence of dopants in the band gap could effectively improve photocatalytic performance through inducing defect states which assist in charge separation of photogenerated electrons and holes [26]. However, all transition metal dopants did not show more perfect photocatalytic activities than iron dopant in the TiO 2 photocatalytic NH 3 synthesis process [54]. Due to their lower Fermi level, transition metals act as electron sinks which creates a Schottky barrier, trapping photogenerated electrons and minimizing the probability for carrier recombination [20]. erefore, reducing carrier recombination is a significant issue to design photocatalytic materials for the NH 3 synthesis from N 2 .

Other Metal Oxide-Based Materials.
In addition to titanium oxide-based materials, other metal oxide-based materials as photocatalyst have also obtained more and more attentions on their behaviours of reducing N 2 to NH 3 in the past several decades. All of these metal oxide-based materials reported have different degrees of photocatalytic activities for NH 3 synthesis from N 2 under ambient conditions. Iron was early used as catalysts in the catalytic ammonia production processes by Dhar and Chowdhry [43]. As we know, iron plays an important role in the Haber-Bosch process, but it is ferric oxide (Fe 2 O 3 ) not iron as the alternative to titanium that could photocatalytic the reduction of N 2 to NH 3 . However, pure Fe 2 O 3 is not capable of photocatalytic NH 3 synthesis from N 2 unless partially reduced [20] and hydrous [23] Fe 2 O 3 . To solve the problem, Khader et al. [35] partially reduced a-Fe 2 O 3 to Fe 3 O 4 for photoactivating N 2 , which resulted in detecting NH 3 in aqueous slurry of the catalyst. Furthermore, the photoactivity of mesoporous β-Ga 2 O 3 nanorods in N 2 photoreduction was ameliorated in the presence of different alcohols, such as ethanol and tert-butanol [32]. Besides ZnO prepared by means of wet etching or precipitation methods, it was reported that the NH 3 yield of unmodified ZnO was higher than that of Pt-loaded ZnO materials [33].
As-synthesized bismuth monoxide (BiO) materials were applied in the photocatalytic reduction of N 2 to NH 3 under solar light. Research results [28] showed that the NH 3 synthesis rate is up to 1226 mmol·g −1 ·h −1 which is about 1000 times higher than that of the Fe-TiO 2 photocatalyst (Table 2). Obviously, deactivation of this photocatalyst did not occur even after being used more than 120 hours. Bismuth oxyhalides, BiOX (X � Cl, Br, and I), have recently become popular due to their excellent applications in photocatalytic NH 3 synthesis from N 2 . BiOBr has also been focused on their material defects. Li et al. [29] employed BiOBr nanosheets with oxygen vacancies (OVs) to reduce N 2 under visible light, and the N 2 reduction rate was estimated to be 104.2 mmol·h −1 .
e structure of BiOBr provides enough space to polarize atoms, and the efficient separation and transformation of charge carriers must depend on the as-formed internal electric field [30]. Moreover, the large number of OVs on the surface of BiOBr hampered the recombination of electron-hole pairs. Interestingly, except the N-type semiconductors above, there are some P-type semiconductors which should be also suitable for the photocatalytic N 2 reductions, such as CrO, MnO, FeO, NiO, Cu 2 O, VO 2 , Cr 2 O 3 , and Ag 2 O. In addition, being added into some impurity atoms such as boron atoms, aluminum atoms, indium atoms, and gallium atoms, the others would turn into P-type semiconductors whose conductivity (i.e., charge carriers) mainly depends on positive vacant holes.
So far, hydrated P-type semiconductors as photocatalysts have been mainly used to reduce N 2 to NH 3 . Early in 1987, hydrous FeO photocatalysts were discovered by Tennakone et al. [55] and were superior to TiO 2 in the case of reducing N 2 to NH 3 under visible light [56]. Hydrous Cu 2 O after impregnation with CuCl was able to photoreduce N 2 to NH 3 . Presumably, the high NH 3 yield resulted from the reduction sites Cu 2 O·xH 2 O [34]. In 1992, vanadium-doped hydrous FeO could enhance reducing N 2 to NH 3 and the average NH 3 production rate was 200 mM·h −1 or so, which benefited from the V impurities increasing vacant holes [57] whose concentration is much larger than the free electrons concentration. erefore, the more the impurity is added, the higher the concentration of multivacant holes, the stronger the conductivity of the semiconductor.

Metal Sulfide-Based Materials.
Similar to the metal oxide-based photocatalysts, metal sulfides have recently become a hot research topic in the field of photocatalytic NH 3 synthesis due to their strong absorption of visible light (Table 3). e NH 3 synthesis rate is up to 325 μmol·g −1 ·h −1 in the presence of ultrathin MoS 2 as photocatalysts, but without charged excitons, the bulk MoS 2 did not have the ability of reduction of N 2 under the same conditions [32]. Miyama et al. [42] suggested that NH 3 yield of CdS/Pt binary photocatalysts was drastically higher than that of pristine CdS under UV irradiation. CdS/Pt/RuO 2 [33] and Pt/CdS-Ag 2 S/RuO 2 [31] photocatalysts were successively applied to reduce N 2 to NH 3 under visible light. e holes in the valence band trapped the electrons that released from RuO 2 . To keep the high photocatalytic activity of CdS for a longer time, maybe some measures must be taken to stop the degradation of CdS to S and Cd 2+ .
Multicomponent metal sulfides with sulphur vacancies, such as Zn 0.1 Sn 0.1 Cd 0.8 S [34] and Mo 0.1 Ni 0.1 Cd 0.8 S [27], could reduce N 2 as photocatalysts under visible light, and the concentration of sulphur vacancies trapping electrons were linear related to the NH 3 yields. In addition, G-C 3 N 4 / ZnSnCdS [42] and g-C 3 N 4 /ZnMoCdS [35] were, respectively, employed for the fixation of N 2 . A tight junction coupling between g-C 3 N 4 and ZnMoCdS was the key for efficient charge transfer.
Only nitrogenases can catalyze reduction of N 2 to NH 3 at room temperature and atmospheric pressure [55]. e nitrogenase complex consists of two proteins: Fe-protein which is responsible for the supply of electrons and MoFe-protein which uses the provided electrons to reduce N 2 to NH 3 [57]. Mimicking the catalytic activity sites FeMocofactor of MoFe-protein, a synthetic complex of Fe, Mo, and S should be taken into account, and organic-sulfide catalysts have also been designed for enhanced N 2 fixation activity. Banergee et al. [39] demonstrated that synthetic FeMoS inorganic clusters could reduce N 2 to NH 3 in aqueous media under light, which showed that structural analogues of nitrogenase can be functional to photocatalytic N 2 fixation. Katherine et al. [31] reported that cadmium sulfide (CdS) nanocrystals can be used to drive the enzymatic reduction of N 2 to NH 3 by photosensitizing the MoFeprotein of nitrogenase not ATP hydrolysis. Under optimal conditions, the turnover rate was 75 per minute, 63% of the ATP-coupled reaction rate for the nitrogenase complex. Both [Mo 2 Fe 6 S 8 (SPh) 3 ] and [Fe 4 S 4 ] clusters comprised in nitrogenase could do so at ambient temperature and pressure, and results suggest [58] that the nitrogenase could keep its photocatalytic activity when the [Fe 4 S 4 ] clusters were replaced with other inert ions such as Sb 3+ , Sn 4+ , and Zn 2+ . In this process, Fe is necessary but Mo is not; however it does not mean that Mo is not playing a role in N 2 binding. In addition, the biohybrids of CdS and MoFe-protein provide a photochemical model for achieving the photocatalytic NH 3 production process. erefore, redox-active iron sulfide containing clusters with high-energy photoexcited states could photocatalyze the reduction of N 2 to NH 3 .

Conclusions
Although the Haber-Bosch process is an important chemical industrial approach to the worldwide population, this traditional NH 3 production process is a high energyconsumption and environmental pollution process. erefore, there have been more and more research interests in photocatalytic reduction of N 2 to NH 3 in the past decades. Difficultly, photocatalytic NH 3 synthesis from N 2 also needs to overcome the energy barrier of N≡N triple band under the spin-polarized plane-wave pseudopotential method. e Density-functional theory has been used to calculate the electronic band structures and the optical absorption spectra of nitrogen-doped and oxygen-deficient anatase TiO 2 . e calculated results are in good agreement with our experimental measurements. ese calculations reveal that the optical absorption of nitrogen-doped TiO 2 in the visible light region is primarily located between 400 and 500 nm, while that of oxygen-deficient TiO 2 is mainly above 500 nm. ese results have important implications for the understanding and further development of photocatalytic materials that are active under visible light. However, synthesis of photocatalytic NH 3 is still kept in a laboratory-scale level.
is work would like to assist in understanding state-ofthe-art in the photocatalytic NH 3 synthesis field for promoting the research field of N 2 fixation. Past reports on photocatalytic NH 3 synthesis focused primarily on titanium oxide-based materials. However, recent studies demonstrated that the N 2 reduction could occur in different photocatalytic reaction systems composed of transition metal-doped materials, metal sulfide-based materials, and other metal oxide-based materials (e.g., P-type semiconductors) as photocatalysts, respectively. Unfortunately, up to our knowledge, no viable and efficient photocatalysts for sustainable NH 3 synthesis could meet all requirements of an active, selective, scalable, long-lived catalyst. With the application of modern computational and experimental techniques, developing efficient photocatalysts should be encouraged to discover the mechanism on how N 2 reduction happens with the photocatalysts in the molecular level.

Conflicts of Interest
e author declares that there are no conflicts of interest regarding the publication of this manuscript.