Recent Advances in Electrocatalysts for Ammonia Oxidation Reaction

: Ammonia (NH 3 ) is a clean energy source that can either be directly used as fuel or a hydrogen carrier due to its high energy density and high hydrogen content. The NH 3 electro-oxidation reaction (AOR) is the main reaction in both direct NH 3 fuel cells and NH 3 electrolysis. The AOR is thermodynamically favorable; however, the sluggish kinetics of the reaction can result in issues such as high overpotential, slow reaction rate, deactivation, etc. To overcome this, multiple strategies have been discussed to develop electrocatalysts that maintain a robust reaction rate in low overpotential regions. In this review, the fundamentals of AOR, including thermodynamics, kinetics, and experimental techniques, are studied. This review also focused on recent progress for catalyst modiﬁcations and their effects, with a particular focus on Pt-or Ni-based electrocatalysts. Additionally, vacant rooms needed to be developed was pointed, and a way to overcome the limitations was suggested. The fundamentals and efforts to prepare catalysts reviewed in this work will be effective in proposing and designing new robust electrocatalysts leading to advance AOR in practice.


Introduction
Ammonia (NH 3 ) is the most potential carbon-free energy carrier due to its remarkable hydrogen storage capacity (17.7 wt%) and energy density (3000 Wh kg −1 ).Additionally, NH 3 can be effectively liquefied at ambient temperature and mild pressure under 8 bar, which makes it superior in terms of storage and transportation compared to other hydrogen carriers [1][2][3].Furthermore, contrary to the traditional Haber-Bosch process, the synthesis of NH 3 using a clean electrochemical approach resulted in a decrease in CO 2 emission, which makes NH 3 a sustainable hydrogen transport media and carbon-free fuel itself [4][5][6][7][8][9].
With increasing attention to NH 3 as a green energy source, various reactions, such as NH 3 oxidation and decomposition, have been actively studied [10][11][12][13][14][15].The use of NH 3 can be mainly divided into two ways; one is to produce electricity by directly feeding it as a fuel like methanol and hydrogen in the fuel cell, and the other is to produce hydrogen via its electrolysis [16,17].Considering the standard reduction potential for the NH 3 oxidation and reduction ( N 2 + 6H 2 O + 6e − → 2NH 3 + 6OH − , E • = −0.77V vs. SHE) relative to that for the hydrogen (2H + + 2e − → H 2 , E • = 0.00 V vs. SHE) required for operating in both energy devices, it has a great benefit not only as a fuel but also as a hydrogen carrier.For example, the theoretical minimal cell potential for hydrogen production in NH 3 electrolysis, requiring 0.06 V, is much smaller than the value in water electrolysis, requiring 1.23 V [18,19].With these considerations, it is no doubt that robust electrochemical direct NH 3 oxidation (AOR, Ammonia oxidation reaction) is critical to use NH 3 in diverse energy conversion devices.Despite the thermodynamic advantage of AOR, its fundamental challenges associated with AOR include sluggish kinetics involving multiple rate-determining steps, unintentional side reactions, and irreversible reactions, such as poisoning [20][21][22].Thus, Notably, NH 3 oxidation reaction (Ammonia oxidation reaction, AOR) seems to be the determining factor in regards to the overall performance of both electrochemical energy conversion devices.Thus, the robust reaction under low overpotential is the most important to develop energy applications using NH 3 .
In the Pourbaix diagram for N 2 -H 2 O in Figure 1a, thermodynamic behavior for AOR depending on pH and potential is discussed [9].The red line represents the theoretical standard reduction potential for the reduction of N 2 , which corresponds to the required potential for the NH 3 oxidation to N 2 and H 2 .In the universal pH range, AOR is more thermodynamically favorable than the oxygen evolution reaction (OER), and thus, the challenges coming from the competitive reaction of OER seem not to be dominant.However, experimental observations have shown significant gaps from this theoretical expectation due to extremely high overpotential requirements.Thus, in fact, experimental AOR can be competitive with OER; furthermore, the produced oxygen can additionally oxygenate undesired nitrogen species of N 2 O, NO, and so on [29,30].These challenges can be expected based on kinetic limitations for AOR.
The reaction steps include 6 electrons releasing accompanied with 2 NH 3 molecules and OH − anions converting to N 2 and 3 H 2 molecules in alkaline media, which have been expected in two different ways, as shown in Figure 1b.In the Oswin-Salomon mechanism, an NH 3 molecule (NH 3,ads ) is adsorbed onto the surface of catalysts and is further oxidized to be an adsorbed N atom (N ads ) with releasing electrons [31].Finally, the two N ads atoms are recombined together to form an N 2 molecule, which subsequently desorbs from the catalyst surface.However, this reaction pathway is limited by two main factors; at low current densities, the rate-determining step is the dehydrogenation reaction from NH 2,ads to NH ads ; at high current densities, the recombination of two N ads becomes the rate-determining step.Additionally, the strongly adsorbed N ads on the catalyst surface can cause catalyst poisoning.
On the other hand, Gerischer and Mauerer proposed another mechanism (Gerischer-Mauerer mechanism), where the first step is the dehydrogenation of NH 3,ads to NH x,ads (x = 1 or 2) intermediates using OH − with releasing electrons [32].The obvious difference between this mechanism and the Oswin-Salomon mechanism is that the two dehydrogenated NH x,ads molecules are dimerized and then further dehydrogenated into N 2,ads .Generally, dimerization is the most representative sluggish step for AOR [20,21].Experimental reports especially using Pt catalysts, mainly support AOR mechanisms using the Gerischer-Mauerer suggestions [1,20,33].
Meanwhile, in order to directly interpret experimental observations, including intermediates and by-product formation and catalyst deactivation, recently, extended AOR mechanisms have been proposed based on theoretical calculation and experimental observations, as shown in Figure 1c [34].The intermediate of N 2 H x+y proposed in the Gerischer-Mauerer mechanism for N 2 production was experimentally revealed by in-situ/operando measurements using FTIR, ATR-SEIRAS, and online electrochemical MS [2,35,36].Although the pathway of NO x formation, causing catalyst poisoning, has not been clearly identified yet, their formation during AOR under high overpotential was also experimentally revealed using in-situ techniques [37].
Besides the theoretical fundamentals of AORs, it is also important to use basic experimental techniques to evaluate the performance of the reaction.The electrochemical activity, e.g., the current density, is investigated using cyclic voltammetry (CV) with a conventional three-electrode configuration (using a working electrode, a reference electrode, and a counter electrode).Figure 1d represents a CV plot obtained using a Pt working electrode measured in 1 M KOH electrolyte with NH 4 + (red line) and without NH 4 + (black line).In the case of electrolytes without NH 4 + , peaks are observed attributing to hydrogen desorption in 0.05-0.45V RHE and Pt surface oxidation from 0.45 V RHE while anodic scanning.Notably, peaks representing hydrogen adsorption and Pt oxide reduction were observed during the cathodic scanning [38,39].Interestingly, in the case of electrolytes with NH 4 + , a distinct oxidation peak is observed at 0.45 V RHE and the maximum current density is shown at 0.75 V RHE , which suggests NH 3 oxidation has occurred [1].The observed peak at 0.2 V RHE during cathodic scanning represents the desorption of NH 3 and H 2 .
The actual amounts of NH 3 can be quantified using diverse detection tools, such as UVvis spectrometer, ion chromatography, and so on.Typically, a colorimetric method using the indophenol blue indicator has been widely suggested [40,41].The indicator comprising hypochlorite, phenolate, and ferricyanide catalyst turns the solution from yellow to green due to the Berthelot reaction with NH 4 + in the solution.The collected adsorption data at 655 nm are shown in Figure 1e,f.Based on the standard calibration curve drawn using the known concentrations of NH 4 + , the amount of NH 4 + presented in the solution can be determined.In addition, N 2 produced by AOR is also quantitatively investigated using the gas chromatography [42].The by-products can be quantified using the Watt and Chrisp method for N 2 H 4 and Griess method for the NO x species [43][44][45].
To determine the performance of AOR, the major three efficiencies must be determined using the quantified results as discussed above [40,42].The first one is NH 3 removal efficiency (R%), which corresponds to the conversion of NH 3 to nitrogen-containing compounds as described in Equation ( 3), where [NH 3 ] 0 and [NH 3 ] t are the initial concentration of NH 3 added and remaining NH 3 in the electrolyte after electrolysis, respectively.The NH 3 removal efficiency indicates the conversion of NH 3 to all products containing N elements, for example, N 2 , NO 2 − , NO 3 − , and so on.
Another important factor is the selectivity for the NH 3 to N 2 conversion as described in Equation ( 4 − able to be produced from AOR, respectively. The last one is the faradaic efficiency, which represents the performance to release electrons from NH 3 -N 2 conversion against the total amount of released electrons, is given in Equation (5).Q theoretically released e − , indicates the overall amounts of generated electrons during potential constant electrolysis, as observed in the J-t plot, and the other Q, actually released e − , represents the number of electrons generated from the only NH 3 -N 2 conversion.In addition, the constant of 6 indicates that 6 electrons can be generated from two NH 3 molecules to produce one N 2 molecule.
So far, fundamentals, challenges, experimental investigation methods, and efficiencies to determine AOR were overviewed.Considering what is discussed above, to improve the performance, strategies to design effective catalysts are highly required.In the next section, promising strategies to prepare effective electrocatalysts and how each strategy affects AOR's performances will be reviewed.

Electrocatalysts for Ammonia Oxidation
This section mainly investigates two groups of NH3 oxidation catalysts and their performances: (1) Pt−based electrocatalysts and (2) Ni−based electrocatalysts.

Pure Pt Electrocatalysts
Noble metals (Ru, Rh, Pd, Pt, and Ir) and transition metals (Au, Ag, and Cu) have been considered to be used as catalysts in AOR.Ru, Rh, and Pd typically have too strong of an affinity for Nads to produce N2, whereas transition metals have relatively too weak NH3 adsorption to dehydrogenate NH3 [46].However, Pt is satisfactory in terms of NH3 adsorption and dehydrogenation capacity, NHx,ads dimerization and product desorption; for this reason, Pt has been considered the most active catalyst and widely studied to be used in AOR [20].To develop the catalytic performance of Pt with reducing the disadvantage of high cost, multiple strategies have been proposed, such as morphology and shape modification, exposure facet engineering, and so on [18,47,48].In this part, recent efforts are mainly reviewed about the way to increase surface area and to provide catalytically active facets of Pt (100).Furthermore, it is discussed how each suggestion affects AOR performance.

Electrocatalysts for Ammonia Oxidation
This section mainly investigates two groups of NH 3 oxidation catalysts and their performances: (1) Pt-based electrocatalysts and (2) Ni-based electrocatalysts.

Pure Pt Electrocatalysts
Noble metals (Ru, Rh, Pd, Pt, and Ir) and transition metals (Au, Ag, and Cu) have been considered to be used as catalysts in AOR.Ru, Rh, and Pd typically have too strong of an affinity for N ads to produce N 2 , whereas transition metals have relatively too weak NH 3 adsorption to dehydrogenate NH 3 [46].However, Pt is satisfactory in terms of NH 3 adsorption and dehydrogenation capacity, NH x,ads dimerization and product desorption; for this reason, Pt has been considered the most active catalyst and widely studied to be used in AOR [20].To develop the catalytic performance of Pt with reducing the disadvantage of high cost, multiple strategies have been proposed, such as morphology and shape modification, exposure facet engineering, and so on [18,47,48].In this part, recent efforts are mainly reviewed about the way to increase surface area and to provide catalytically active facets of Pt (100).Furthermore, it is discussed how each suggestion affects AOR performance.
Guntae Kim et al. synthesized two different types of Pt electrodes, which have uneven dense morphology or flower-shape depending on electrocatalytically synthetic condition [42].The unevenly dense shaped Pt (CP-Pt) electrode was prepared using chronopotentiometry (CP) at a constant current of 2 mA, whereas the flower-shaped Pt (X CV-Pt, where X is the number of CV cycles; 100, 200, 300, 400, 500 and 600) electrode was deposited using cyclic voltammetry (CV) in the potential range of −0.4 to 0.2 V Ag/AgCl , as shown in Figure 2a.Unlike during the CP process, H 2 bubbles are generated via water reduction while scanning potential negatively than hydrogen reduction potential during the CV process.The evolution and detachment of the H 2 bubble at a particular Pt site induce a flower-shaped Pt formation.The AOR activity of the Pt−NCs was evaluated using CV with 0.1 M NH4OH in 1 M KOH and was compared with that of commercial Pt nanoparticles (Pt−NCs−JM) purchased from Johnson Matthey Corporation.In order to remove any possibility of affecting AOR coming from the different surface areas of Pt samples, the AOR activities were normalized using the collected ECSA value, as shown in Figure 3c.For both Pt samples, a remarkable peak indicating AOR starts at 0.5 VRHE and the highest current density was observed at 0.7 VRHE.Pt−NCs exhibited a peak current density of 5.1 mA cmECSA −2 , whereas Pt−NCs−JM exhibited a peak current density of 2.6 mA cmECSA −2 .Given that Pt−NCs−JM comprises polycrystalline Pt, which has diverse Pt facets, it can be inferred that the advanced AOR performance in Pt−NCs is mainly due to the dominant Pt (100) facets in the sample.
Catalytic stability during electrolysis is another important factor in determining AOR performance.Generally, the strong adsorption of nitrogen atoms (Nads) on Pt catalysts leads to active site deactivation, resulting in severe instability.The catalytic stability of the Pt electrodes was evaluated using constant potential electrolysis at 0.6 VRHE with the same electrolyte conditions.From Figure 3d, the time−profiled current density over the mass of the loaded Pt catalysts (J−t) showed that Pt−NCs exhibited a higher initial current density and lower current density decay rate relative to Pt−NCs−JM.The Pt (100) facets are able to improve initial catalytic activity and stability for AOR by providing an appropriate binding strength for the intermediate.
Juan M. Feliu et al. also prepared (100) facet−dominant cubic platinum nanoparticles (Pt NPs) using the water−in−oil microemulsion method [51].The Pt (100) facet fraction was adjusted by controlling the concentration of HCl added during synthesis.As the concentration of HCl is increased from 0% to 25%, the (100) fraction also increases from 10% To investigate the effect of morphology on AOR activity, the CV process was conducted with 1 M NH 3 in 5 M KOH.From Figure 2b, the current density began to increase at −0.65 V Ag/AgCl representing AOR and exhibited a maximum current density around −0.3 V Ag/AgCl ; the current density then decreased due to strong adsorption of N ads .The CV-Pt electrodes exhibited a higher peak current density than the CP-Pt electrodes; notably, the 600 CV-Pt electrode recorded a maximum current density of 560 mA cm −2 , which is three times higher than that of the CP-Pt electrode, as shown in Figure 2b,c.Since all electrodes have similar polycrystalline crystal structures and crystallinity of Pt as investigated using X-ray diffraction (XRD), it can be excluded the possibility of increasing the activity caused by a preferable crystal structure or facet for AOR.Instead, the major contribution to the increased current density is the enhanced active surface area for flower-shaped CV-Pt compared to unevenly dense structured CP-Pt.This work clearly reveals that the morphology, as well as the surface area of the electrocatalyst, is one of the critical factors affecting AOR performances.
W.B. Hu et al. prepared Pt particles with different morphology on indium tin oxide film (ITO) substrate by electrodeposition with controlling the current density from 0.2 mA cm −2 to 1 mA cm −2 [49].With an increased current density, the Pt growth rate at the edge of the nucleus for Pt is faster than the rate at the other sites of the nucleus, resulting in a morphological change from spherical to flower-like.
To compare AOR activity between different electrodes, CVs were measured using Pt electrodes with spherical and flower-like morphologies with 0.1 M NH 3 in 1 M KOH.To prevent false evaluation due to irregular loading amounts of Pt on the electrode, the obtained CV data were normalized with the loaded catalyst amounts.The flower-like Pt particles exhibited a peak current density of 0.063 mA µg −1 , which is three times higher than that of the spherical Pt particle.It is mainly due to the difference in the surface area accompanied by each morphology.In fact, the flower-shaped Pt electrode exhibited two times higher electrochemical active surface area (ECSA) than the Pt spherical Pt electrode.Interestingly, when comparing AOR performance divided into ECSA of each Pt electrode, still, the flower-shaped Pt electrode showed advanced activity compared to spherical Pt.Besides the large surface area, the tips and sharp-edged flakes comprising flower morphology develop additional surface electric fields, affecting the enhancement of AOR kinetics.
Facet engineering is another important area to explore in terms of improving AOR kinetics.Yu Chen et al. synthesized Pt nanocubes (Pt-NCs) with dominant (100) facets using a facile hydrothermal method involving the agents of polyvinyl pyrrolidone (PVP), 2-methylimidazole, and ascorbic acid [50].Each agent is critical in the preparation of the desired catalyst; for example, PVP served as the stabilizing agent, 2-methylimidazole was used as (100) facet-selective agent, and ascorbic acid was used to form nanocubes by controlling the reducing rate of Pt.The preferential (100) facet in the Pt-NCs was revealed by TEM images, as shown in Figure 3a,b.
The AOR activity of the Pt-NCs was evaluated using CV with 0.1 M NH 4 OH in 1 M KOH and was compared with that of commercial Pt nanoparticles (Pt-NCs-JM) purchased from Johnson Matthey Corporation.In order to remove any possibility of affecting AOR coming from the different surface areas of Pt samples, the AOR activities were normalized using the collected ECSA value, as shown in Figure 3c.For both Pt samples, a remarkable peak indicating AOR starts at 0.5 V RHE and the highest current density was observed at 0.7 V RHE .Pt-NCs exhibited a peak current density of 5.1 mA cm ECSA −2 , whereas Pt-NCs-JM exhibited a peak current density of 2.6 mA cm ECSA −2 .Given that Pt-NCs-JM comprises polycrystalline Pt, which has diverse Pt facets, it can be inferred that the advanced AOR performance in Pt-NCs is mainly due to the dominant Pt (100) facets in the sample.
Catalytic stability during electrolysis is another important factor in determining AOR performance.Generally, the strong adsorption of nitrogen atoms (N ads ) on Pt catalysts leads to active site deactivation, resulting in severe instability.The catalytic stability of the Pt electrodes was evaluated using constant potential electrolysis at 0.6 V RHE with the same electrolyte conditions.From Figure 3d, the time-profiled current density over the mass of the loaded Pt catalysts (J-t) showed that Pt-NCs exhibited a higher initial current density and lower current density decay rate relative to Pt-NCs-JM.The Pt (100) facets are able to improve initial catalytic activity and stability for AOR by providing an appropriate binding strength for the intermediate.
Juan M. Feliu et al. also prepared (100) facet-dominant cubic platinum nanoparticles (Pt NPs) using the water-in-oil microemulsion method [51].The Pt (100) facet fraction was adjusted by controlling the concentration of HCl added during synthesis.As the concentration of HCl is increased from 0% to 25%, the (100) fraction also increases from 10% to 44%; when the concentration of HCl is increased beyond 25%, the (100) fraction begins to decrease.To investigate how the (100) facet fraction affects AOR performance, CV was measured with 0.1 M NH 3 in 0.2 M NaOH.The Pt sample with a (100) facet fraction of 44% exhibited a current density six times higher than that of the sample with a (100) fraction of 10%.This work obviously suggested a way to increase desired Pt (100) fraction via adjusting the concentration of HCl agents and, furthermore, demonstrated the effect of the Pt (100) fraction on the AOR performance.
The promising effect of Pt (100) facet engineering was theoretically supported by density functional theory (DFT) studies [52,53].Mavrikakis et al. compared the onset potential and activation barrier of Pt (100) to Pt (111) for each Gerischer-Mauerer and Oswin-Salomon mechanism, as shown in Figure 3e [54].For the Gerischer-Mauerer mechanism, on Pt (100), the PDS is the deprotonation of HNNH ads to NNH ads with the onset potential of 0.29 V, which is similar to that of 0.28 V on Pt (111).However, NH 2,ads dimerization to N 2 H 4, has a higher barrier of 2.32 eV on Pt (100) than the activation barrier of 1.07 eV on Pt (111).It represents the ability of HNNH ads deprotonation on Pt (100) is close to on Pt (111).However, N 2 H 4, formation by NH 2,ads dimerization on Pt (111) is more favorable than the dimerization on Pt (100).On the other hand, based on the Oswin-Salomon mechanism on Pt (100), the potential determining step (PDS) is the deprotonation of NH 2, ads to NH ads with the onset potential of 0.87 V, and the N ads dimerization to N 2 has a low activation barrier of 0.53 eV.On the other hand, on Pt (111), the PDS (NH ads to N ads ) has an onset potential of 0.52 V, which is lower than that of Pt (100), and the N ads dimerization has a high activation barrier of 2.24 eV, which is much higher than that of Pt (100).It indicates that N ads formation on Pt (100) is more difficult than the intermediate formation on Pt (111) referred by onset potentials.However, once N ads formed, its dimerization and desorption on Pt (100) is more preferred than that on Pt (111) when comparing the activation barrier.Considering the theoretically calculated data, Pt (100) facet can provide the lowest activation barrier for N 2 formation following the Oswin-Salomon mechanism and also can explain experimental advances in Pt (100) engineering.
44% exhibited a current density six times higher than that of the sample with a (100) fraction of 10%.This work obviously suggested a way to increase desired Pt (100) fraction via adjusting the concentration of HCl agents and, furthermore, demonstrated the effect of the Pt (100) fraction on the AOR performance.
The promising effect of Pt (100) facet engineering was theoretically supported by density functional theory (DFT) studies [52,53].Mavrikakis et al. compared the onset potential and activation barrier of Pt (100) to Pt (111) for each Gerischer−Mauerer and Oswin−Salomon mechanism, as shown in Figure 3e [54].For the Gerischer-Mauerer mechanism, on Pt (100), the PDS is the deprotonation of HNNHads to NNHads with the onset potential of 0.29 V, which is similar to that of 0.28 V on Pt (111).However, NH2,ads dimerization to N2H4, has a higher barrier of 2.32 eV on Pt (100) than the activation barrier of 1.07 eV on Pt (111).It represents the ability of HNNHads deprotonation on Pt (100) is close to on Pt (111).However, N2H4, formation by NH2,ads dimerization on Pt (111) is more favorable than the dimerization on Pt (100).On the other hand, based on the Oswin−Salomon mechanism on Pt (100), the potential determining step (PDS) is the deprotonation of NH2, ads to NHads with the onset potential of 0.87 V, and the Nads dimerization to N2 has a low activation barrier of 0.53 eV.On the other hand, on Pt (111), the PDS (NHads to Nads) has an onset potential of 0.52 V, which is lower than that of Pt (100), and the Nads dimerization has a high activation barrier of 2.24 eV, which is much higher than that of Pt (100).It indicates that Nads formation on Pt (100) is more difficult than the intermediate formation on Pt (111) referred by onset potentials.However, once Nads formed, its dimerization and desorption on Pt (100) is more preferred than that on Pt (111) when comparing the activation barrier.Considering the theoretically calculated data, Pt (100) facet can provide the lowest activation barrier for N2 formation following the Oswin−Salomon mechanism and also can explain experimental advances in Pt (100) engineering.

Pt-Ir Bimetal Electrocatalysts
Pt is considered the most active AOR catalyst among noble metals.However, it should overcome the catalytic deactivation caused by the strong adsorption of N ads ; this behavior is called surface poisoning.Pt-based binary and ternary catalysts with Ir, Ru, Rh, and Pd have been suggested to reduce surface poisoning and improve AOR performance [23,[55][56][57][58].
Among these, Ir is the most representative candidate for Pt-based binary and ternary catalysts because it has a sufficient NH 3 dehydrogenation capacity and appropriately low N ads affinity.For example, Pt-Ir bimetal catalysts can decrease potential onset relative to a single Pt catalyst and reduce surface poisoning by altering the electron state of Pt and weakening the adsorption strength for N ads [59,60].In this part, recent studies on Pt-Ir alloy synthesized in various ways and their origins for advancing AOR activity are discussed.
Minhua Shao et al. synthesized a catalyst by depositing Ir onto Pt (100) facets on carbon substrate (Ir-decorated Pt NCs/C) using the modified solvothermal method [59].The uniform distribution of Ir on Pt was revealed by Energy-dispersive X-ray spectroscopy (EDS) mapping images in Figure 4a.The effect of Ir on Pt was evaluated by comparing LSVs between Ir-decorated Pt NCs/C and pure Pt NCs/C, in which current density was normalized using ECSA in the electrolyte of 0.1 M NH 3 in 0.1 M KOH.In Figure 4b, although both Ir-decorated Pt NCs/C and Pt/NCs exhibited pre-peaks at 0.30 V RHE, indicating NH 3 adsorption, the current density of Ir-decorated Pt NCs/C arose from a much lower onset potential of 0.43 V RHE than that of Pt NCs/C started at 0.55 V RHE .Furthermore, the Irdecorated Pt NCs/C exhibited a current density of 1.26 mA cm −2 , which was two times higher than that of Pt NCs/C; this confirms that Ir decorated on Pt results in not only reduced overpotential but also increased AOR kinetics.Pt−based ternary metal alloying is another promising strategy for advancing AOR performance.Particularly, the addition of a third transition metal onto Pt−Ir alloy has been mainly studied.Gan Wu et al. prepared a ternary PtIrNi alloy on supports comprising SiO2 and CNT−COOH using the sonochemical−assisted method, as shown in Figure 5a [61].The supports can provide a large surface area for the deposition of the three metals.
The AOR activities of various electrodes, including Pt/SiO2−CNT−COOH, Besides advancing in kinetic, catalytic stability was also improved by Ir decoration on Pt.The current density decay rates of the Ir-decorated Pt NCs/C and Pt NCs/C were compared using constant potential electrolysis at 0.6 V RHE .The current density of the Irdecorated Pt NCs/C decayed in 100 s, whereas that of the Pt NCs/C decayed immediately.Considering the major origin of instability observed for Pt electrocatalysts, it reveals that Ir decoration on Pt is effective in preventing Pt surface poisoning by storing N ads .
To further investigate the role of Ir in this regard, the differences in Gibbs free energies of the Pt (100) and Ir-decorated Pt (100) were monitored using DFT calculations for each step of the AOR pathways, as shown in Figure 4c.The Ir-decorated Pt (100) exhibited a relatively low activation energy barrier of 0.74 eV for the dominant rate-determining step compared to pure Pt (100), which exhibited an energy barrier of 0.96 eV.From the experimental and theoretical investigations, it is found that the critical role of Ir on Pt is to improve reaction kinetics by lowering the energy barriers and to increase stability by preventing Pt poisoning.
Elena A. Baravona et al. synthesized Pt-Ir alloys using the polyol method under a pH of 7 (named PtIr (1)) and 8.5 (named PtIr (2)) [60].As shown in Figure 4d, the AOR activity for both Pt-Ir alloys and the single Pt was evaluated using CV with 0.5 M NH 4 OH in 1 M KOH, where the measured current densities were normalized using the value of ECSA of each sample.In terms of peak current density, PtIr (1) was less effective than pure Pt; however, PtIr (1) had a clear advantage in onset potential compared to pure Pt.PtIr (1) showed an onset potential for AOR at −1.07 V MSE , which was 0.13 V less than that of the single Pt.The results indicate that the PtIr (1) alloy is better at facilitating high catalytic AOR activity in the low potential region than the single Pt.
Furthermore, the catalytic stabilities of Pt and PtIr (1) were evaluated using a timeprofiled potential constant electrolysis at −0.8 V MSE , as shown in Figure 4e.The PtIr (1) showed a higher initial current density and slower current density decay than that of the single Pt.These advances in terms of onset potential, the current density at low potential, and stability can mainly be attributed to the electronic effect between Pt and Ir atoms in the PtIr (1) alloy.Meanwhile, PtIr (2) exhibited an even worse AOR performance than pure Pt in terms of kinetics and stability; this is due to the formation of Pt oxide on the surface of the electrode during synthesis, unlike PtIr (1) and Pt which contain metallic Pt state.In the XPS spectra of Pt 4f, as shown in Figure 4f, PtIr (2) showed additional peaks at 72.9 eV and 74.4 eV, which corresponds to the oxidized Pt, while that of the metallic Pt is at 71.6 eV.The Pt oxide on PtIr (2) inhibited the adsorption of NH 3 , resulting in a high onset potential and low stability even as alloying Ir with Pt.
Pt-based ternary metal alloying is another promising strategy for advancing AOR performance.Particularly, the addition of a third transition metal onto Pt-Ir alloy has been mainly studied.Gan Wu et al. prepared a ternary PtIrNi alloy on supports comprising SiO 2 and CNT-COOH using the sonochemical-assisted method, as shown in Figure 5a [61].The supports can provide a large surface area for the deposition of the three metals.
The AOR activities of various electrodes, including Pt/SiO 2 -CNT-COOH, PtIr/SiO 2 -CNT-COOH, and PtIrNi-SiO 2 -CNT-COOH, were compared using CVs with 0.1 M NH 3 in 1.0 M KOH; their current densities collected during CV were normalized by preciousgroup-metal (PGM) mass, as shown in Figure 5b.PtIrNi/SiO 2 -CNT-COOH exhibited the lowest onset potential, followed by PtIr/SiO 2 -CNT-COOH and Pt/SiO 2 -CNT-COOH; this confirms that the presence of Ni in the alloy reduces the onset potential.For example, PtIrNi/SiO 2 -CNT-COOH exhibited the lowest onset potential at 0.4 V RHE and the highest current density of 124.0A g −1 .PtIr/SiO 2 -CNT-COOH and Pt/SiO 2 -CNT-COOH even recorded more positive values of onset potential, as much as 0.05 V RHE and 0.15 V RHE , respectively, compared to that from PtIrNi/SiO 2 -CNT-COOH.Moreover, the Ni ternary alloy exhibited the highest stability, which was measured at a constant 0.65 V RHE , among all Pt-based electrodes, as shown in Figure 5c.
alloy is a promising approach to preparing electrocatalysts for AOR in terms of onset potential and stability.However, even if AOR performance has been largely developed by multi−modification using Pt, for example, morphology control, facet engineering, alloy, and so on, the high cost of Pt and Pt−based electrodes remain a significant challenge for advancing AOR in practice.Therefore, the rational design of electrocatalysts comprising non−precious elements is highly required.We summarized the AOR performance on Pt−based catalysts in Table 1.The enhancement of AOR performance by alloying with Pt, Ir and Ni was elucidated by DFT calculations.According to Gibbs's free energy changes for each elemental step of AOR, as shown in Figure 5d, the Pt-Ir alloy had a largely reduced thermodynamic barrier for dehydrogenation compared to the single Pt.Moreover, as alloying Ni with Pt-Ir, the barrier was further reduced significantly.By adding Ni to Pt-Ir alloy, d orbital states can be largely increased, resulting in facilitating NH ads formation and adsorption, which is the major rate-determining step in the single Pt.This work demonstrated that a ternary metal alloy is a promising approach to preparing electrocatalysts for AOR in terms of onset potential and stability.However, even if AOR performance has been largely developed by multi-modification using Pt, for example, morphology control, facet engineering, alloy, and so on, the high cost of Pt and Pt-based electrodes remain a significant challenge for advancing AOR in practice.Therefore, the rational design of electrocatalysts comprising non-precious elements is highly required.We summarized the AOR performance on Pt-based catalysts in Table 1.

Ni-Based Metal Electrocatalysts
Pt and Pt-based noble metals have been considered active catalysts in AOR because of their low overpotential and high N 2 selectivity; however, their high cost and susceptibility to deactivation have proven to be significant challenges.Recently, non-noble metal-based electrocatalysts have been proposed as an alternative to Pt-based electrocatalysts [76][77][78].
Ni has been studied as a potential catalyst to replace noble metals because it is intrinsically satisfactory for adsorbed NH 3 dehydrogenation and NH x,ads dimerization and shows experimental activity in AOR with its abundancy [20,40,76].Notably, Ni exhibited significantly promising long-term stability for a few hours, while Pt-based catalysts were deactivated within a few seconds.However, formation of by-products, N 2 H 4 , NO 2 − , NO 3 − , and so on, leads to decrease selectivity for N 2 production [79].To overcome the limitation and further improve AOR activity, Ni-based binary and ternary catalysts have been studied [80][81][82].This section focuses on recent research regarding combinations of Ni with hetero-atoms and their effects on AOR.
Ni alloying with Cu elements has been widely proposed to promote AOR performance by overcoming the challenges of imperfect reaction rate and N 2 selectivity [19].
Shanwen Tao et al. synthesized Ni-Cu layered hydroxides (Ni 0.8 Cu 0.2 LHs) on carbon fiber cloth using the hydrothermal method, followed by electrochemical activation over 200 CV cycles in the potential range of −0.2 V Ag/AgCl to 1.0 V Ag/AgCl , as shown in Figure 6a  TBA + ions were exchanged into Ni 2+ and Cu 2+ ions.Additionally, uniformly electrodeposited surface was identified through the SEM images as shown in Figure 7b,c.The electrocatalytic AOR performance for NiCu/MnO2 was evaluated using CV with 55 mM NH4Cl in 0.5 M NaOH, as shown in Figure 7d.NiCu/MnO2 exhibited a peak at 0.53 VHg/HgO, confirming the oxidation from Ni 2+ to Ni 3+ , which was followed by a rapid increase in current density, an indication of facile AOR.Although Ni/MnO2 also exhibited two distinct increases in current density, attributing to Ni 2+ oxidation at 0.65 VHg/HgO and AOR, the subsequent increase in current density was much lower than that of NiCu/MnO2.These results obviously represented that the addition of Cu 2+ into Ni 2+ intercalated MnO2 reduced the onset potential for Ni 2+ oxidation and increased the current density for AOR.Moreover, the NiCu/MnO2 showed promising catalytic stability during constant potential electrolysis at 0.6 VHg/HgO for 24 h, as shown in Figure 7e.Even if the current density steadily decreased as time passed, the current density was recovered in the 2nd−run, in a re−prepared fresh electrolyte after 24 h.This indicates that the current density decay was not originated from deactivation or electrode corrosion but from the rapid consumption of NH3 at the NiCu/MnO2 electrode.In addition, using online gas chromatography, the desired N2 product selectivity was estimated to be 97.4%.This work revealed that the addition of Cu 2+ in Ni 2+ intercalated in MnO2 could enhance AOR performance by mediating Furthermore, the catalytic stability of Ni 0.8 Cu 0.2 LHs was evaluated using constant potential electrolysis of 0.55 V Ag/AgCl under the same electrolytic conditions discussed above.As shown in Figure 6d, the addition of NH 4 Cl resulted in an increase in current density for both Ni 0.8 Cu 0.2 LHs and bare Ni(OH) 2 electrodes.Interestingly, Ni 0.8 Cu 0.2 LHs showed a stable current density of 35 mA cm −2 , which was approximately seven times higher than that of bare Ni(OH) 2 for at least 3 h.The NH 3 removal efficiency, determined at 0.55 V Ag/AgCl , reached 84%, whereas bare Ni(OH) 2 recorded an NH 3 removal efficiency of 58%.Based on all the results discussed above, the Cu-doped Ni LDH was able to enhance AOR performance in terms of activity, stability, and ultimate NH 3 removal efficiency.
Masaharu Nakayama et al. prepared Ni 2+ and Cu 2+ co-intercalated electrocatalysts on layered MnO 2 (NiCu/MnO 2 ) using a two-step synthetic process: the electrodeposition of tetrabutylammonium cations (TBA + )/MnO 2 at a constant potential of 1.0 V Ag/AgCl and then the ion exchange reaction, carried out by immersing the catalyst in a solution containing Ni 2+ and Cu 2+ [26].As shown in Figure 7a, TBA/MnO 2 showed a peak at 7.14 • , which represents a (001) plane of layered MnO 2 and a d-spacing of 1.23 nm.On the other hand, in NiCu/MnO 2 , the peak of the (001) plane split and appeared at 10 • and 12 • , and the d-spacing decreased to 0.88 nm and 0.75 nm.The reduced d-spacing indicates that the bulky TBA + ions were exchanged into Ni 2+ and Cu 2+ ions.Additionally, uniformly electrodeposited surface was identified through the SEM images as shown in Figure 7b,c.Ni−Cu−Fe oxyhydroxides onto carbon paper (NiCuFe electrode) using electrodeposition at a constant potential of −1.15 VHg/Hg2Cl2, followed by electrochemical activation [27].The AOR activity of the NiCuFe electrode was evaluated using CV with 55 mM NH4Cl in 0.5 M NaOH and compared with that of the NiCu electrode.The NiCuFe electrode exhibited a rapid increase in current density over 0.36 V Hg/Hg2Cl2, and its current density reached 60 mA cm −2 at 0.7 V Hg/Hg2Cl2, whereas the current density of the NiCu electrode also increased over 0.45 V Hg/Hg2Cl2 and its current density only reached 30 mA cm −2 at 0.7 V Hg/Hg2Cl2.This shows that the addition of Fe to NiCu significantly reduced the onset potential for AOR and increased the current density generated by AOR.Moreover, in the stability test during the constant potential electrolysis at 0.55 V Hg/Hg2Cl2, the NiCuFe electrode exhibited an initial current density of 40 mA cm −2 , which was two times higher than that of the NiCu electrode.The current density decreased as NH3 was consumed.The NiCuFe electrode recorded a high NH3 removal efficiency of 95 % quantified using the indophenol method, and a selective NH3 conversion to N2 also occurred, with a minor product of NO3 − .The electrocatalytic AOR performance for NiCu/MnO 2 was evaluated using CV with 55 mM NH 4 Cl in 0.5 M NaOH, as shown in Figure 7d.NiCu/MnO 2 exhibited a peak at 0.53 V Hg/HgO , confirming the oxidation from Ni 2+ to Ni 3+ , which was followed by a rapid increase in current density, an indication of facile AOR.Although Ni/MnO 2 also exhibited two distinct increases in current density, attributing to Ni 2+ oxidation at 0.65 V Hg/HgO and AOR, the subsequent increase in current density was much lower than that of NiCu/MnO 2 .These results obviously represented that the addition of Cu 2+ into Ni 2+ intercalated MnO 2 reduced the onset potential for Ni 2+ oxidation and increased the current density for AOR.
Moreover, the NiCu/MnO 2 showed promising catalytic stability during constant potential electrolysis at 0.6 V Hg/HgO for 24 h, as shown in Figure 7e.Even if the current density steadily decreased as time passed, the current density was recovered in the 2nd-run, in a re-prepared fresh electrolyte after 24 h.This indicates that the current density decay was not originated from deactivation or electrode corrosion but from the rapid consumption of NH 3 at the NiCu/MnO 2 electrode.In addition, using online gas chromatography, the desired N 2 product selectivity was estimated to be 97.4%.This work revealed that the addition of Cu 2+ in Ni 2+ intercalated in MnO 2 could enhance AOR performance by mediating the electron transfer between Ni 2+ and NH 3 and facilitating the oxidation of adsorbed NH 3 .
Lately, to further develop AOR performance, researchers have suggested adding the third hetero-metal to Ni-Cu-based electrocatalysts.Junmin Xue et al. synthesized triple Ni-Cu-Fe oxyhydroxides onto carbon paper (NiCuFe electrode) using electrodeposition at a constant potential of −1.15 V Hg/Hg2Cl2 , followed by electrochemical activation [27].The AOR activity of the NiCuFe electrode was evaluated using CV with 55 mM NH 4 Cl in 0.5 M NaOH and compared with that of the NiCu electrode.The NiCuFe electrode exhibited a rapid increase in current density over 0.36 V Hg/Hg2Cl2, and its current density reached 60 mA cm −2 at 0.7 V Hg/Hg2Cl2 , whereas the current density of the NiCu electrode also increased over 0.45 V Hg/Hg2Cl2 and its current density only reached 30 mA cm −2 at 0.7 V Hg/Hg2Cl2 .This shows that the addition of Fe to NiCu significantly reduced the onset potential for AOR and increased the current density generated by AOR.Moreover, in the stability test during the constant potential electrolysis at 0.55 V Hg/Hg2Cl2 , the NiCuFe electrode exhibited an initial current density of 40 mA cm −2 , which was two times higher than that of the NiCu electrode.The current density decreased as NH 3 was consumed.The NiCuFe electrode recorded a high NH 3 removal efficiency of 95 % quantified using the indophenol method, and a selective NH 3 conversion to N 2 also occurred, with a minor product of NO 3 − .The effect of doping Fe into NiCu electrode on AOR activity was additionally revealed using DFT calculation.According to DFT calculation, the first step of the reaction, the adsorption of NH 3 on an oxygen atom bonded to Ni and the NiCu electrode, showed the highest energy barrier of 4.91 eV among the tailored elementary steps.The addition of Fe, which has relatively low electronegativity, to the NiCu electrode, resulted in a significant increase in electron density around oxygen atoms and lowered the energy barrier to 1.08 eV.This suggests that Ni-Cu alloys doped with low electronegative elements, such as Fe, are able to accelerate the adsorption of NH 3 and ultimately improve AOR kinetics.
Besides Ni and Cu combination, diverse mixture of hetero-atoms and each element's effects have been experimentally proposed for AOR.Jianhui Wang et al. prepared physically mixed Ag and Ni electrocatalysts (Ag/Ni) by facile hand-mixing method [83].The Ag/Ni was measured AOR performances by CV with 0.5 M NH 3 in 1.5 M NaOH and compared its activity with pristine Ag and Ni, respectively.The pristine Ag represented no catalytic activity in the potential region of CVs, whereas the pristine Ni exhibited a current density increase from 0.6 V Hg/HgO .However, its dominant product from electrolysis was not N 2 but O 2 , indicating OER is dominant rather than AOR on the electrode having Ni alone.On the other hand, the Ag/Ni showed a significant current density increase from lower onset potential at 0.46 V Hg/HgO , which is a much higher current density than pristine Ni.Furthermore, Ag/Ni produced N 2 as a major product, representing its superior activity for AOR.The faradaic efficiency for N 2 production was recorded at 59.8%, while constant potential electrolysis at 0.7 V Hg/HgO for 2 h.Furthermore, from the control experiments of CVs with and without NH 3, as shown in Figure 8a, the synergistic effect of Ag and Ni for AOR was discussed.As anodic scanning with NH 3 , the peak for Ag + oxidation to Ag 2+ , typically observed during the scan without NH 3 , disappeared and which indicates the oxidized Ag + can form a strong Ag + -NH 3 complex, suppressing further NH 3 oxidation.On the other hand, pristine Ni exhibited similar CV results regardless of NH 3 presence in the solution because the competitive reaction of OER was more favorable than AOR on the Ni-only surface.Interestingly, Ag/Ni facilitated each Ag + -NH 3 complex formation and OH − adsorption on Ni, and finally, they synergistically promoted to produce N 2 by applying potentials over 0.4 V Hg/HgO, as shown in Figure 8b.
lower onset potential at 0.46 VHg/HgO, which is a much higher current density than pristine Ni.Furthermore, Ag/Ni produced N2 as a major product, representing its superior activity for AOR.The faradaic efficiency for N2 production was recorded at 59.8%, while constant potential electrolysis at 0.7 VHg/HgO for 2 h.Furthermore, from the control experiments of CVs with and without NH3, as shown in Figure 8a, the synergistic effect of Ag and Ni for AOR was discussed.As anodic scanning with NH3, the peak for Ag + oxidation to Ag 2+ , typically observed during the scan without NH3, disappeared and which indicates the oxidized Ag + can form a strong Ag + -NH3 complex, suppressing further NH3 oxidation.On the other hand, pristine Ni exhibited similar CV results regardless of NH3 presence in the solution because the competitive reaction of OER was more favorable than AOR on the Ni−only surface.Interestingly, Ag/Ni facilitated each Ag + -NH3 complex formation and OH − adsorption on Ni, and finally, they synergistically promoted to produce N2 by applying potentials over 0.4 VHg/HgO, as shown in Figure 8b.The studies for developing AOR using Pt and Ni−based electrocatalysts have been mainly reported, whereas the research using other elemental compositions has been rarely studied.Furthermore, although new groups of catalysts, for example, CuO, CuSn(OH)6, TiO and so on, were suggested, their AOR performances still need to be advanced [77,84,85].Their recorded AOR activities are summarized in Table 2.It is highly required not only to search for new latent electrocatalyst candidates but also to advance their AOR performance using effective strategies.The studies for developing AOR using Pt and Ni-based electrocatalysts have been mainly reported, whereas the research using other elemental compositions has been rarely studied.Furthermore, although new groups of catalysts, for example, CuO, CuSn(OH) 6 , TiO and so on, were suggested, their AOR performances still need to be advanced [77,84,85].Their recorded AOR activities are summarized in Table 2.It is highly required not only to search for new latent electrocatalyst candidates but also to advance their AOR performance using effective strategies.

Summary and Perspective
NH 3 has been gaining attention as a sustainable fuel and a carbon-free hydrogen carrier with a high energy density and high hydrogen content.Given this, the electrochemical NH 3 oxidation (AOR) is a key reaction for generating electricity in NH 3 fuel cells and producing H 2 fuel in NH 3 electrolyzers.However, the sluggish kinetics of AOR give rise to severe drawbacks of overpotential and reaction rate, even though it is thermodynamically favorable.These experimental challenges have been overcome by developing modified electrocatalysts, as reviewed in this work.
Pt electrocatalysts have been considered for use in AOR due to their superior NH 3 affinity and rapid AOR kinetics; because of this, Pt electrocatalysts have been further studied by exploring the diverse strategies modifying morphology and exposure facet on the surface, alloying, and so on.Particularly, a Pt-Ir was able to significantly improve the reaction kinetics of AOR by effectively reducing the energy barrier.Meanwhile, recently, cost-effective non-noble metal catalysts, especially Ni-based catalysts, have also been intensively suggested as alternatives to Pt.In particular, Ni-Cu bimetal catalysts exhibited the most promising AOR performance in terms of reaction rate and stability by adjusting the affinity of NH 3 and intermediates on Ni and Cu.
However, still searching for robust new catalysts and their development is necessary in order to perform AOR in practice.Especially to develop cost-effective non-noble metalbased electrocatalysts for AOR, several challenges need to be overcome.First, selective AOR in low overpotential regions is important.However, typical non-noble metal-based electrocatalysts exhibited AOR performance in high overpotential, and thus, the AOR competes with competitive reactions, such as water oxidation or NH 3 oxidation to NO 3 − or NO 2 − , and so on.However, it was found that forming alloys with hetero-atoms can alter the electronic structure of catalysts resulting in reducing the NH 3 adsorption energy barrier and effectively reducing the overpotential [27,61,89].
Another important factor is AOR rate, which can be determined by current density.Doping or alloying can also be effective in increasing the current density by stabilizing reaction intermediates and enhancing reaction kinetics.In addition, the engineering morphology of electrocatalysts can be a key suggestion to facilitate reactions.The nano-sized electrocatalyst provides lots of active catalytic sites in large surface areas, leading to advanced current densities for AOR.Moreover, combining electrocatalysts with conductive supports or substrates comprising electrodes can promote significantly improved AOR performance.Substrates, such as carbon paper, carbon nanotube (CNT), Ni foam, and so on, have large surface areas and high intrinsic conductivities, reducing electron transfer resistance within the electrocatalysts coupled with the substrates and improving the AOR activity [61,88,91].
Unlike catalyst modifications, it has been rarely studied evaluating the total performance, including AOR and the counter reduction reaction in single-cell and engineering environmental conditions.For example, in order to effectively produce energy from NH 3 fuel or to extract H 2 from NH 3 electrolysis in practice, it is necessary to construct a singlecell typed real device composed of the developed catalysts and evaluate performance using the device.To our knowledge, the best faradaic efficiency of N 2 and H 2 production was 90% and 90%, respectively, from NH 3 during current density constant electrolysis at 200 mA cm −2 , where the potential difference of 0.85 V was measured in the single-cell composed with two Pt electrodes [42].In addition, the total performance can be improved by optimizing reaction environmental conditions.For instance, the reaction operating temperature can be a critical factor directly affecting to the reaction rate [16,[92][93][94][95].As referring to the previous work, the device's current density was largely increased to 250 mA cm −2 at 60 • C compared to the current density of 80 mA cm −2 measured at room temperature.Therefore, to use NH 3 as a promising energy source efficiently, it is highly necessary to engineer and optimize device set-up and operating conditions.
Besides developing catalysts, a fundamental understanding of the reaction mechanism is critical to further improve AOR.Currently, the AOR mechanism has been mainly investi-gated as suitable for Pt catalysts.However, actual reaction mechanisms or rate-determining steps can be different depending on the types of electrocatalysts.The DFT calculations have been widely applied to propose reaction mechanisms and elucidate the key effect for the catalytic modification [20,96,97].With advances in theoretical investigation, observations of reaction mechanisms experimentally are important by monitoring adsorbed intermediates.Recently, efforts to investigate actual reaction mechanisms on electrocatalysts have been reported using in-situ or operando characterizations [1,21,98].Considering the progress, the AOR has significant potential to develop as practically by combining catalytic modification and device engineering and fundamental studies.

Figure 1 .
Figure 1.(a) Pourbaix diagram for the N2−H2O system.Reproduced with permission [9].Copyright 2022, MDPI.(b) Schematic diagram of the NH3 oxidation pathways.(c) The extended AOR mechanism includes NOx species.The O−S mechanism indicates the Oswin−Salomon mechanism, and the G−M mechanism indicates the Gerischer−Mauerer mechanism.* indicates adsorption on the catalyst surface.Reproduced with permission [34].Copyright 2022, Elsevier B. V. (d) Cyclic voltammograms of Pt plate in the presence (red line) and absence (black line) of 0.1 M NH3 in 1 M KOH with a scanning rate of 20 mV s -1 .(e) UV−Vis absorption spectra of indophenol assays with various NH3 concentrations.(f) Calibration curve used for the NH3 quantification.Reproduced with permission [41].Copyright 2023, Elsevier B. V.

Figure 1 .
Figure 1.(a) Pourbaix diagram for the N 2 -H 2 O system.Reproduced with permission [9].Copyright 2022, MDPI.(b) Schematic diagram of the NH 3 oxidation pathways.(c) The extended AOR mechanism includes NO x species.The O-S mechanism indicates the Oswin-Salomon mechanism, and the G-M mechanism indicates the Gerischer-Mauerer mechanism.* indicates adsorption on the catalyst surface.Reproduced with permission [34].Copyright 2022, Elsevier B. V. (d) Cyclic voltammograms of Pt plate in the presence (red line) and absence (black line) of 0.1 M NH 3 in 1 M KOH with a scanning rate of 20 mV s -1 .(e) UV-Vis absorption spectra of indophenol assays with various NH 3 concentrations.(f) Calibration curve used for the NH 3 quantification.Reproduced with permission [41].Copyright 2023, Elsevier B. V.

Catalysts 2023, 13 , x 7 of 23 Figure 2 .
Figure 2. (a) Schematic illustration of the CP−Pt and CV−Pt electrodes.(b) CVs results and (c) Comparison of peak current density and current density at −0.4 VAg/AgCl for CV−Pts and CP−Pts.All electrocatalytic measurements were conducted with 1 M NH3 in 5 M KOH.Reproduced with permission [42].Copyright 2021, Royal Society of Chemistry.

Figure 2 .
Figure 2. (a) Schematic illustration of the CP-Pt and CV-Pt electrodes.(b) CVs results and (c) Comparison of peak current density and current density at −0.4 V Ag/AgCl for CV-Pts and CP-Pts.All electrocatalytic measurements were conducted with 1 M NH 3 in 5 M KOH.Reproduced with permission [42].Copyright 2021, Royal Society of Chemistry.

Figure 3 .Figure 3 .
Figure 3. (a) The low and (b) the high magnified TEM images of Pt−NCs.(c) the CV curves normalized using the value of ESCA with a potential scan rate of 5 mV s −1 and (d) the time−profiled current density over the mass loaded on the electrode (J−t curves) at 0.6 VRHE for Pt−NCs and Pt−NPs−JM.All AOR performances were measured with 0.1 M NH4OH in 1 M KOH.Reproduced with permission [50].Copyright 2020, Elsevier B. V. (e) Free−energy diagram for AOR on Pt (100) at 0.0 V.In purple are NHx,ads (x = 0 or 1 or 2) dimerization elementary steps with the respective transition-state Figure 3. (a) The low and (b) the high magnified TEM images of Pt-NCs.(c) the CV curves normalized using the value of ESCA with a potential scan rate of 5 mV s −1 and (d) the time-profiled current density over the mass loaded on the electrode (J-t curves) at 0.6 V RHE for Pt-NCs and Pt-NPs-JM.All AOR performances were measured with 0.1 M NH 4 OH in 1 M KOH.Reproduced with permission [50].Copyright 2020, Elsevier B. V. (e) Free-energy diagram for AOR on Pt (100) at 0.0 V.In purple are NH x,ads (x = 0 or 1 or 2) dimerization elementary steps with the respective transitionstate (TS) energies.N + N mechanism indicates the Oswin-Salomon mechanism, and * indicates adsorption on the catalyst surface.Reproduced with permission [54].Copyright 2021, Elsevier B. V.

Catalysts 2023 ,
13,  x 10 of 23 the single Pt.These advances in terms of onset potential, the current density at low potential, and stability can mainly be attributed to the electronic effect between Pt and Ir atoms in the PtIr (1) alloy.Meanwhile, PtIr (2) exhibited an even worse AOR performance than pure Pt in terms of kinetics and stability; this is due to the formation of Pt oxide on the surface of the electrode during synthesis, unlike PtIr (1) and Pt which contain metallic Pt state.In the XPS spectra of Pt 4f, as shown in Figure4f, PtIr(2) showed additional peaks at 72.9 eV and 74.4 eV, which corresponds to the oxidized Pt, while that of the metallic Pt is at 71.6 eV.The Pt oxide on PtIr (2) inhibited the adsorption of NH3, resulting in a high onset potential and low stability even as alloying Ir with Pt.

Figure 4 .
Figure 4. (a) HAADF−STEM images and corresponding EDS mapping images of Ir−decorated Pt NC.(b) the LSV curves were measured with 0.1 M NH3 in 0.1 M KOH, where the current densities were normalized using the value of ECSA.(c) Gibbs free energy changes for each step of AOR pathways on Pt(100) and Ir−decorated Pt(100).Reproduced with permission [59].Copyright 2020, Elsevier B. V. (d) The CV curves measured with 0.5 M NH4OH in 1 M KOH, where current densities were normalized using the value of ECSA, (e) J−t curves at −0.8 VMSE of Pt, PtIr (1) and PtIr (2) and (f) Pt 4f XPS spectra for PtIr (1) and PtIr (2).Reproduced with permission [60].Copyright 2013, Elsevier B. V.

Figure 4 .
Figure 4. (a) HAADF-STEM images and corresponding EDS mapping images of Ir-decorated Pt NC.(b) the LSV curves were measured with 0.1 M NH 3 in 0.1 M KOH, where the current densities were normalized using the value of ECSA.(c) Gibbs free energy changes for each step of AOR pathways on Pt (100) and Ir-decorated Pt (100).Reproduced with permission [59].Copyright 2020, Elsevier B. V. (d) The CV curves measured with 0.5 M NH 4 OH in 1 M KOH, where current densities were normalized using the value of ECSA, (e) J-t curves at −0.8 V MSE of Pt, PtIr (1) and PtIr (2) and (f) Pt 4f XPS spectra for PtIr (1) and PtIr (2).Reproduced with permission [60].Copyright 2013, Elsevier B. V.
[24].The AOR performance of Ni 0.8 Cu 0.2 LHs was evaluated using CV with 55 mM NH 4 Cl in 0.5 M NaOH and compared with that of bare Ni(OH) 2 .In Figure6b, both electrodes exhibited an increase in the current density at 0.43 V Ag/AgCl , indicating the oxidation from hydroxide to oxyhydroxide, such as the transformation of Ni 0.8 Cu 0.2 OOH from Ni 0.8 Cu 0.2 LHs (Ni 0.8 Cu 0.2 (OH) 2 ) and NiOOH from Ni(OH) 2 , respectively.In the potential over 0.43 V Ag/AgCl representing the AOR region, Ni 0.8 Cu 0.2 LHs showed a rapid increase, and it represents a higher current density than that from Ni(OH) 2 .Additionally, the Tafel slope of Ni 0.8 Cu 0.2 LHs was much smaller than that of Ni(OH) 2 , indicating the electron transfer within Ni 0.8 Cu 0.2 LHs is relatively fast, as shown in Figure 6c.From this, it can be posited that the addition of Cu to the Ni oxyhydroxide improved both electron transfer and AOR kinetics.The first origin of AOR enhancement for Ni 0.8 Cu 0.2 LHs is a large surface area compared to Ni(OH) 2 .Moreover, the adjusted strength of NH 3 adsorption promotes the formation of the first intermediate of the Ni 1-x Cu x OOH(NH 3 ) ads .Catalysts 2023, 13, x 14 of 23

Figure 6 .
Figure 6.(a) Schematic illustration for the preparation of Ni 0.8 Cu 0.2 LHs on carbon fiber cloth substrate, (b) the CVs and (c) the Tafel plots for Ni 0.8 Cu 0.2 LHs and bare Ni(OH) 2 .(d) the J-t curves at 0.55 V Ag/AgCl .The AOR performance was measured with 55 mM NH 4 Cl in 0.5 M NaOH.Reproduced with permission [24].Copyright 2017, Elsevier B. V.

Figure 7 .
Figure 7. (a) The XRD patterns of TBA/MnO2 (black line), Ni/MnO2 (blue line), Cu/MnO2 (green line), NiCu/MnO2 before AOR (upper red line) and NiCu/MnO2 after AOR (lower red line).(b) The low and (c) the high−magnified SEM images of NiCu/MnO2 deposited on carbon paper.(d) The CV curves and (e) the J−t curves at 0.6 VHg/HgO for NiCu/MnO2, Ni/MnO2, and Cu/MnO2.The AOR performance was measured with 55 mM NH4Cl in 0.5 M NaOH.Reproduced with permission [26].Copyright 2021, American Chemical Society.The effect of doping Fe into NiCu electrode on AOR activity was additionally revealed using DFT calculation.According to DFT calculation, the first step of the reaction, the adsorption of NH3 on an oxygen atom bonded to Ni and the NiCu electrode, showed the highest energy barrier of 4.91 eV among the tailored elementary steps.The addition of Fe, which has relatively low electronegativity, to the NiCu electrode, resulted in a significant increase in electron density around oxygen atoms and lowered the energy barrier to 1.08 eV.This suggests that Ni−Cu alloys doped with low electronegative elements, such as Fe, are able to accelerate the adsorption of NH3 and ultimately improve AOR kinetics.Besides Ni and Cu combination, diverse mixture of hetero−atoms and each elementʹs effects have been experimentally proposed for AOR.Jianhui Wang et al. prepared physically mixed Ag and Ni electrocatalysts (Ag/Ni) by facile hand−mixing method [83].The

Figure 8 .
Figure 8.(a) The CV curves of Ni, Ag and Ag/Ni on the glassy carbon electrode.The AOR performance was measured with and without 0.5 M NH3 in 1.5 M NaOH.(b) Schematic mechanism of Ag and Ni in the AOR.Reproduced with permission [83].Copyright 2022, Royal Society of Chemistry.

Figure 8 .
Figure 8.(a) The CV curves of Ni, Ag and Ag/Ni on the glassy carbon electrode.The AOR performance was measured with and without 0.5 M NH 3 in 1.5 M NaOH.(b) Schematic mechanism of Ag and Ni in the AOR.Reproduced with permission [83].Copyright 2022, Royal Society of Chemistry.

Table 1 .
A brief summary of experimental studies on AOR using various Pt-based electrocatalysts.

Table 2 .
A brief summary of experimental studies on AOR using various non-noble metal electrocatalysts.