Recent progress on electrocatalysts in ammonia electrooxidation reaction for clean hydrogen production

Abstract

Hydrogen (H₂) is considered one of the most promising alternative energy resources, serving as a clean energy carrier, replacing fossil fuels. Ammonia (NH₃) is recognized as a high-potential H₂ carrier and is utilized as a primary material for H₂ production. The NH₃ electro-oxidation reaction (AOR) is crucial for both direct NH₃ fuel cells and NH₃ electrolysis. Nevertheless, AOR frequently encounters difficulties including a high overpotential, sluggish reaction rates, and catalyst deactivation as a result of its slow kinetics, even though it is thermodynamically advantageous. In order to overcome these obstacles, numerous approaches have been suggested for the creation of electrocatalysts that can sustain a strong reaction rate even when exposed to low overpotential. The kinetics, mechanisms, and experimental methods of AOR are the primary focus of this overview. It also examines recent advancements in electrocatalyst modifications, with a particular emphasis on single-atom electrocatalysts and transition metal-based catalysts. Additionally, the paper discusses the techno-economic assessment of the AOR reaction for H₂ production. Furthermore, future perspectives and recommendations for electrocatalyst development in AOR are provided. Proposing and designing new, resilient electrocatalysts will be significantly aided by the fundamentals and efforts detailed in this review; this, in turn, will propel advancements in AOR for practical applications.

---

Introduction

Fossil fuels, notably coal, oil, and natural gas, have historically served as primary energy sources across various sectors, such as electricity generation, transportation, and industrial processes. However, their finite nature, coupled with significant environmental and social externalities, underscores the imperative for transitioning towards renewable and alternative energy sources. Amidst this transition, hydrogen (H₂) has emerged as a versatile energy carrier with the potential to decarbonize multiple sectors of the economy, including industry, transportation, and power generation.^1,2 While H₂ can be produced through diverse pathways, including electrolysis of water using renewable electricity,³ biomass gasification,⁴ and steam reforming of natural gas coupled with carbon capture and storage,⁵ conventional methods often entail carbon-intensive processes or rely on non-renewable resources thus undermining the environmental benefits of H₂ as a clean energy vector.

In this context, the electrooxidation of ammonia (NH₃) presents a promising pathway toward sustainable H₂ production, addressing both environmental and technical challenges associated with conventional methods.^6,7 By harnessing the electrochemical reactivity of NH₃, researchers endeavour to develop efficient and scalable technologies for on-demand H₂ generation, thereby contributing to the transition towards a low-carbon, H₂-based economy. Notably, NH₃ electrooxidation can be conducted under mild conditions, obviating the need for high temperatures and pressures inherent in conventional steam reforming processes.⁸ This condition reduces energy consumption and enhances process safety and operational flexibility.

The progression of publications focusing on the electrooxidation of NH₃ from 2008 to 2024, depicted in Fig. 1, underscores the increasing research interest in this area. Notably, 449 documents, including articles, review articles, proceeding papers, and book chapters, have comprehensively explored various aspects of NH₃ electrooxidation over the past 16 years. The surge in publications since 2010 is indicative of significant advancements in research techniques and the perceived potential applications of NH₃ electrooxidation in energy conversion, environmental remediation, and chemical synthesis. This surge is also reflected in the increase in citations by 2023. Furthermore, the overlay visualization in Fig. 2 highlights key terms such as “electrooxidation” and “electrochemical oxidation,” which stand out with the most significant node sizes, indicating their high frequency of occurrence. Additionally, terms represented by yellow nodes, such as catalysts, electrocatalysts, oxidation, mechanism, and fuel cell, underscore the primary focus of current research.

Fig. 1 Publication trends in the area of electrooxidation of NH₃ within 2008–2024.

Fig. 2 Overlay visualization of most frequent terms attained via VOSViewer v.1.6.19.

Indeed, designing electrocatalysts is pivotal in enhancing the efficiency and selectivity of NH₃ electrooxidation. Various materials and strategies have been developed to produce efficient catalysts to replace platinum (Pt),⁹ which is costly and scarce. Researchers have explored transition metal-based catalysts, such as nickel (Ni),^10,11 cobalt (Co),¹² and iron (Fe),¹³ as promising alternatives due to their abundance and lower cost compared to Pt. Additionally, alloying different metals or incorporating dopants into these materials has been investigated to enhance their catalytic activity and stability. Furthermore, non-metallic materials, such as carbon-based¹⁴ and metal nitrides¹⁵ catalysts have also shown potential for AOR. These materials offer advantages such as high surface area, tunable electronic properties, and chemical stability, making them attractive candidates for electrocatalysis.^16,17 Moreover, nanostructuring and surface engineering techniques have been employed to optimize the morphology and active sites of catalyst materials, further improving their performance in catalysing NH₃ electrooxidation.

Although many reviews cover catalyst development, this review will focus specifically on recent progress in electrocatalyst development to enhance AOR performance. The review will explore the various materials and strategies explored for electrocatalyst design, including single-atom electrocatalysts and transition metal-based catalysts. Additionally, the review will include a techno-economic analysis, assessing the feasibility and cost-effectiveness of different production methods of H₂ for large-scale implementation. Furthermore, the review will discuss the future perspectives and recommendations in electrocatalyst development for AOR. Overall, this review aims to provide insights into the state-of-the-art electrocatalyst design for improving the efficiency and selectivity of NH₃ electrooxidation, contributing to the advancement of sustainable H₂ production technologies.

Ammonia (NH₃) electrolysis for green hydrogen (H₂) production

The utilization of NH₃ as a H₂ carrier undoubtedly is one of the exceptional approaches for the in situ production of H₂ fuel. This approach is also noted to exhibit minimal greenhouse gaseous (GHGs) or any hazardous compounds. Besides, NH₃ also displays a promising candidate material for large-scale H₂ production attributed to its high stability and significant capacity for H₂ storage.¹⁸ Concerning these intrinsic features, the production of H₂ from NH₃ has attracted considerable attention and has been deeply studied for its practicability, especially in the industrial process. Interestingly, the transformation of NH₃ to H₂ can be realized using two distinct pathways which are thermal decomposition and electrolysis approaches. Even though the thermal decomposition method is well-known for being more economical compared to electrolysis, the efficiency of this method yet is still inferior to that of the electrolysis method.¹⁹ Moreover, using the high temperature for the decomposition of NH₃ to H₂ and nitrogen constituents will be vulnerable to the formation of GHGs such as nitrous compounds. On the other hand, the electrolysis of NH₃ provides a promising interest in the field of H₂ production. This method utilizes a green technology with no GHGs emissions and it also can achieve high efficiency when powered by renewable energy. The efficiency of this approach also can be further enhanced depending on the advancement of the technology used. Furthermore, thermodynamic calculations demonstrated that the conversion of NH₃ to nitrogen and H₂ requires a negligible quantity of external energy (0.06 V), which is significantly less than the energy demanded by water electrolysis (1.23 V).²⁰ Theoretically, this indicates that the NH₃ to H₂ conversion is more energy-efficient than water-splitting.²¹

Generally, the electrolysis of water using renewable electricity and the electrooxidation of ammonia are two distinct processes with different objectives, methodologies, and challenges. The primary purpose of water electrolysis is to produce H₂ and O₂ gases.²² This process involves splitting water into its constituent elements using an electrical current, with the reaction: 2H₂O → 2H₂ + O₂. The H₂ produced can be used as a clean fuel for various applications, including fuel cells, chemical synthesis, and energy storage. Typically, noble metal catalysts such as platinum (Pt), iridium (Ir), and ruthenium (Ru) are used in this process due to their high activity and stability. The process can utilize either acidic or alkaline electrolytes, and it usually operates at elevated temperatures and pressures to enhance efficiency and gas production rates.²³ The efficiency of water electrolysis systems ranges from 60% to 80%, depending on the specific technology and operating conditions. However, the process requires a high energy input, and challenges include catalyst degradation and overpotential losses. Despite these challenges, water electrolysis is highly suitable for integration with renewable energy sources like solar and wind power, allowing for the storage of intermittent renewable energy in the form of H₂ gas. The environmental impact is minimal, as water is the only by-product, making it a sustainable and environmentally friendly process.

In contrast, the primary purpose of NH₃ electrooxidation is to convert NH₃ to N₂ gas and water helping to treat ammonia-containing waste and reduce environmental pollution. This process typically uses transition metal catalysts such as Pt, Ni, and Cu, which facilitate the oxidation of NH₃. It generally operates at ambient temperature and pressure, making it less energy-intensive than water electrolysis. The efficiency of NH₃ electrooxidation depends on the catalyst used and the reaction conditions. Significant challenges include NH₃ slip (unreacted ammonia), the formation of unwanted by-products, and electrode poisoning. However, this process converts harmful NH₃ into N₂ and water, thereby reducing environmental pollution and preventing ammonia emissions into the atmosphere. NH₃ electrooxidation can also be integrated with renewable energy sources, particularly where NH₃ is produced from renewable H₂, creating a closed-loop sustainable process. Its applications include wastewater treatment and nitrogen removal from industrial effluents, with potential uses in H₂ production.

Briefly, NH₃ electrolysis is a process of dissociating the liquid NH₃ or ammonium in electrochemical cells. Fig. 3 illustrates the basic concept of NH₃ electrolysis. Unlike the photoelectrochemical water-splitting approach, NH₃ electrolysis is a process that is been conducted in the absence of light. This process begins with the direct supply of electricity from the power supply to the system. Typically, in an aqueous NH₃ electrolysis system, as the charges flow through the circuit, the anode electrode undergoes the oxidation of NH₃ to the nitrates compound which then further oxidizes to N₂ molecules. Concurrently, the produced electrons at the anode will be transferred to the cathode electrode for subsequent reduction of water molecules to the H₂ products.²⁴ On this side, the H₂ evolution reaction (HER) will take place by reducing the water molecules with a consumption of six electron species. Moreover, the mechanism of NH₃ electrolysis can be described by the equations below:²⁵

2NH₃ + 6OH⁻ → N₂ + 6H₂O + 6e⁻ (1)

6H₂O + 6e⁻ → 3H₂ + 6OH⁻ (2)

Fig. 3 Principle of electrolysis NH₃ in aqueous solution for H₂ production.

Adsorption of NH₃ onto the surface of the catalyst initiates the six-electron in AOR process as well as takes place in an alkaline environment. Significantly, the AOR is traversed via two discrete pathways (Fig. 4).²⁶ Oswin and Salomon introduced the O–S mechanism, a consistent dehydrogenation process, in 1963 as reported by Kim et al.²⁷ In this mechanism, *NH₃ on the surface of catalyst undergoes sequential dehydrogenation by OH⁻ to form *N, which later combines to form N₂. Thus, the O–S mechanism can be described as shown in eqn (3)–(7):

NH_3(aq.) → *NH₃ (3)

*NH₃ + OH⁻ → *NH₂ + H₂O + e⁻ (4)

*NH₂ + OH⁻ → *NH + H₂O + e⁻ (5)

*NH + OH⁻ → *N + H₂O + e⁻ (6)

*2N → N₂ (7)

Fig. 4 Oxidation of NH₃ by O–S (a) and G–M (b) mechanism. Reproduced from ref. 26 with permission from Elsevier, copyright 2024.

Subsequently to this research, the O–S mechanism was later fine-tuned by Mauerer and Gerischer. Contradict, those researchers proposed that the G–M mechanism in the AOR involves the recombination of *N_xH_y species as well as OH⁻ dehydrogenation.²⁸ The G–M mechanism is represented by eqn (8) and (9) below, where x = 1 or 2, and y = 1 or 2.

*NH_x + *NH_y → *N₂H_x+y (8)

*N₂H_x+y + (x + y)OH⁻ → N₂ + (x + y)H₂O + (x + y)e⁻ (9)

The dehydrogenated *NH and *NH₂ were combined to form *N₂H_x+y. Subsequent dehydrogenation of *N₂H_x+y leads to the formation of N₂. In both major mechanisms, the rate-determining step is the dimerization of *NH_x and *N (N_ads), which significantly limits the production of N₂ from NH₃ oxidation. However, N_ads can be further oxidized by OH⁻ to generate N–O intermediates, which then form nitrite and nitrate.

N_ads + OH⁻ → HON_ads + e⁻ (10)

HON_ads + OH⁻ → HONO_ads + e⁻ (11)

HONO_ads + OH⁻ → ONO_ads + H₂O + e⁻ (12)

ONO_ads + OH⁻ → ONO⁻ (aq.) (13)

ONO_ads + OH⁻ → ONOOH_ads + e⁻ (14)

Driven by the catalyst, the OH_ads produced in the electrolyte possess high bonding energies, which halts the dimerization process of N_ads, leading to their further oxidation to NO₂⁻ and NO₃⁻.²⁶ During this process, under the influence of OH⁻ in the electrolyte, the N_ads adsorbed on the catalyst surface are gradually hydroxylated to form species such as HON_ads, ONO_ads, etc., which are then converted to ONOads before being smoothly oxidized to NO₂⁻ and NO₃⁻. For example, the strongly adsorbed N_ads hinder the further adsorption of NH_x, resulting in a poor AOR reaction rate. A catalyst, such as CuNi, is expected to exhibit favorable AOR activity for N₂ production only if it possesses appropriate adsorption bonding energy for each intermediate (Fig. 5a and b).²⁹

Fig. 5 Ammonia oxidation for the preparation of (a) NO₂⁻ and (b) NO₃⁻ mechanism. Reproduced from ref. 26 with permission from Elsevier, copyright 2024.

Most research supports the idea that intermediates recombine in the AOR. Additionally, analyses like DEMS and SERS have indicated that N₂ cannot be generated into fully dehydrogenated *N; instead, it acts merely as a poisoning species. Several investigations have linked AOR intermediates to negative health effects, which has sparked discussion over AOR intermediates like *N_x, *N_xH_y, and *NO_x. Accordingly, the key to getting high AOR activity is minimizing poisoning by N_x, NO_x, and N_xH_y and improving the recombination of N_xH_y species.

A straightforward and effective defect-engineering strategy offers a robust approach for AOR mechanistic studies, where oxygen vacancies (Vo) are found to effectively modulate the electronic properties of surface CuO and enhance the binding strength of reaction intermediates.³⁰ Conventional Pt-based electrocatalysts are generally believed to follow the G–M mechanism because the key intermediate of N₂H₄ was detected. However, the O–S mechanism reaction route via dehydrogenation of NH₃ to N and N–N coupling is applied to the rich Vo–CuO. Based on the free energy pathway of the AOR process via DFT simulations, the NH₃ adsorption process is energetically unfavourable on the Vo-free CuO surface. The free energy changes of the three dehydrogenation steps are all uphill and higher than 1.4 eV due to weak adsorption towards *NH_x species, with the largest free energy change being 1.75 eV, indicating inferior AOR activity (Fig. 6A).³⁰ However, after introducing the Vo, the free energy of each step decreases by 0.34, 1.84, 2.84, and 2.64 eV, respectively. Additionally, the electronic structures of Cu atoms on the different surface CuO showed that the local density of states of surface Cu atoms move significantly upward towards the Fermi level after introducing Vo, indicating stronger adsorption strength towards the adsorbates. Therefore, the presence of the Vo is beneficial to efficient NH₃ oxidation by stabilizing reaction intermediates. Similarly, Zhang et al. claimed that the AOR pathways follow the N + N mechanism containing the N₂H₃ and N₂H₄ intermediate by using TiO–Vo.³² The introduction of Vo made the N₂ desorption a spontaneous exoergic process, which was a good solution to the most critical catalyst poisoning problem in the AOR process.

Fig. 6 Free energy diagrams of NH₃ oxidation into N₂ on the (A) Vo-rich and Vo-free CuO via O–S mechanism. Reproduced from ref. 30 with permission from Springer, copyright 2022 and (B) CeO₂ and CeO₂–Vo follow the G–M mechanism. Reproduced from ref. 31 with permission from ACS Publications, copyright 2024.

Differently, Yang et al. reported that CeO₂ with Vo species are the active species in the CeO_x/Y electrode during ammonia electrolysis and follow the G–M mechanism involving the N₂H₄ intermediates (Fig. 6B).³¹ Generally, the dehydrogenation of *NH₃ to *NH₂ is usually the rate-determining step for AOR over metal oxides. In this study, the energy barrier for the conversion of *NH₃ to *NH₂ on CeO₂ is 2.13 eV, whereas on CeO₂–Vo it is only −0.26 eV, indicating that NH₃ is readily deprotonated. This shows that the NH₃ oxidation process is energetically unfavourable on the CeO₂ surface, but the presence of Vo in CeO₂ significantly reduces the energy barrier for *NH₃ to *NH₂, aligning with the enhanced AOR activity of CeO₂–Vo. When NH₃ adsorbs on Vo sites, one elongated N–H bond points toward a neighboring H atom. The elongation of the N–H bond and the attraction of neighboring O atoms facilitate H transfer. Therefore, it can be concluded that the presence of Vo in CeO₂ not only enhances NH₃ adsorption but also lowers the energy barrier for NH₃ dehydrogenation. This contributes to the increased activity of NH₃ electrolysis following the G–M mechanism.

Achieving high NH₃ electrolysis performance is the main goal for all researchers. Similar to the water-splitting reaction, optimizing the electrochemical cell of NH₃ electrolysis should take into consideration the type, pH and concentration of the electrolyte, and the electrode materials. Frankly speaking, ammonium-based electrolytes such as NH₄Cl, NH₄Br, NH₄I, NH₄NO₃, and NH₄PF₆ are the most suitable electrolytes for NH₃ electrolysis since this process involves the reduction of ammonium species to generate H₂ gas.³³ In addition, the influence of pH on AOR is something that needs to be taken into consideration. Typically, the current density measurement in the AOR rises steadily as the pH goes up, while the starting potential of the anode decreases in line with pH increments. In the study of AOR, alkaline-based electrolytes remain the most commonly used because NH₃ can stay in its molecular form more effectively in these conditions, generally leading to better AOR performance.²⁶ On the other hand, NH₃ indirect oxidation in acidic media is a very slow process because NH₄⁺ is pushed away from the electrodes in acidic ions. Moreover, the oxidation of NH₃ in low pH media often made the anode material susceptible to undesirable corrosion which greatly hampers the AOR process. However, an area of potential interest for AOR is the use of non-aqueous electrolytes, which can eliminate undesirable water oxidation side reactions, narrow AOR potential apertures, and the formation of by-products like nitrates. Therefore, regulating the pH of an AOR system logically is a crucial factor that merits investigation.

Moreover, the concentration of NH₃ in the electrolyte solution also greatly influences the electrolysis performance. Utilising an electrolyte containing a high concentration of NH₃ in this instance can enhance the interaction between NH₃ and the catalyst, resulting in substantial production of H₂. In a recent study, Tran et al. observed an enhancement in faradaic efficiency along with an increase in NH₃ concentration.³⁴ Similarly, Jo et al. also found that the AOR performance is directly proportional to the initial concentration of NH₃ which concluded that a high concentration of NH₃ is favorable for high H₂ production.³⁵ Despite these circumstances, the employment of a high concentration of NH₃ in the electrolysis process might upsurge the overall process cost and unsuitable for large-scale implementation. Therefore, optimization of economic cost and electrolysis performance are substantially important for the systematic design of NH₃ electrolysis systems.

Furthermore, it has been validated that the electrochemical reaction occurring at the anodic electrode constitutes the rate-limiting step in the NH₃-to-H₂ conversion process. Hence, the selection of an appropriate anodic electrode exhibiting outstanding physicochemical and electrical characteristics is imperative. Notably, there exist specific prerequisites to be fulfilled when opting for the anodic material. Primarily, the material must possess elevated physicochemical stability, ensuring minimal or negligible degradation throughout the electrolysis procedure, thereby preserving its original attributes. Additionally, the stability of the anodic material plays a pivotal role in preventing undesired side reactions that could pose environmental hazards and compromise the efficacy of the anodic electrode in the AOR.³⁶ Moreover, since the anode electrode undergoes an oxidation reaction, the anode must demonstrate exceptional corrosion resistance to ensure an efficient AOR process. Long-term stability of the anodic material during electrolysis is essential for its viability in commercial applications, particularly in industrial processes. Specifically, a high surface area of anode material is desired for the AOR process. The high surface area with exceptional active sites can enhance the surface electrode–electrolyte interface and accelerate the AOR process, ultimately leading to high production of H₂.²⁶ Electronic properties are of paramount importance, with the anodic material requiring excellent electron and ion transport capabilities to sustain high current densities and minimize energy losses during the process. Lastly, to ensure heightened and efficient AOR activity, the anodic material should exhibit low overpotential values and high selectivity towards N₂ production over NO_x compounds.³⁴

Over the years, due to the extraordinary properties of Pt, such as a lower *N affinity than other noble metals and a superior dehydrogenation capacity of NH_x in comparison to non-noble metals, Pt-based materials have garnered considerable interest in the AOR study. This has resulted in superior AOR performance. Nevertheless, the high cost of utilizing Pt has restricted its practical applications. On the other hand, even though non-noble metal demonstrated low AOR efficiency compared to noble metal, its overall efficiency can still be improved by further modification. Therefore, single-atom catalysts and transition-metal-based catalysts have been widely employed for NH₃ electrolysis studies.

Electrocatalysts for AOR in NH₃-to-H₂ conversion

The performance of two classes of AOR catalysts is the primary focus of this section: (1) transition metal-based electrocatalysts and (2) single-atom (SACs) electrocatalysts. Because of their high selectivity, stability, and catalytic activity, SACs electrocatalysts have emerged as promising candidates for NH₃ oxidation. Dispersed on a support material are isolated metal atoms that constitute these catalysts. SACs are distinguished by their electronic and geometric configurations, which enable them to facilitate charge transfer with high efficiency and furnish active sites for the oxidation of NH₃. Recent research has demonstrated the excellent performance of SACs in NH₃ electrooxidation, offering high conversion efficiency and superior selectivity for H₂ production. Additionally, SACs exhibit remarkable stability, making them attractive candidates for practical applications.

Transition metal-based electrocatalysts have also shown promising performance in NH₃ electrooxidation. These electrocatalysts typically consist of transition metal nanoparticles or alloys supported on conductive substrates. Transition metals such as nickel (Ni), cobalt (Co), and copper (Cu) have been extensively studied for NH₃ oxidation due to their high catalytic activity and stability. Transition metal-based electrocatalysts offer several advantages, including tunable electronic properties and high surface area, which enhance catalytic activity and facilitate NH₃ oxidation.⁶ Moreover, the abundance and low cost of transition metals make them attractive candidates for large-scale H₂ production via AOR. Table 1 shows the summary of single atoms catalysts and transition metal-based electrocatalysts for H₂ production via AOR.

Table 1 Single atoms catalyst and transition metal-based electrocatalysts for H₂ production via AOR

Catalyst	Current density (mA cm^−2)	Onset potential (VRHE)	Ref.
Ni films	0.45	0.3	33
Pt nanoparticle	1.45	0.9	24
Pt nanosphere	2.6	—	37
Pt/Rh	2.5	0.53	38
Pt/NC	2.4	0.1	39
PtIrCu nanodendrites	122.9 A g^−1	0.35	40
PtIrCu nanoframe	—	0.33	41
PtRu/C	0.1	0.2	42
PtIrZn	0.0043	0.3	43
Pt/NiO	0.24	−0.48	44
Pt/TiO2/S-OLCN	0.28	−0.501	45
Pt/SnO2/C	—	0.07	14
TM2@g-CN	—	−0.48	46
RuO2–ZnO/Al2O3	0.74	0.2	47
TiO2 nanofibers	0.06	0.4	32
NiO–TiO2	2.93	1.24	10
Co/Ni–C	2.0	0.45	48
NiCO2N nanoneedle	10	0.71	21
NiCo2N nanosheets	10	0.55	15
NiCuBO/NF	25	1.301	49
B-NiFe-LDH/NF	100	1.481	50
Ni(OH)2–Cu2O@CuO	10	0.47	51
Ni1−xCuxOOH	35	0.28	52
Cu2O/NF	4.6	0.55	53
Ni3S4@NiCo2O4/NF	50	0.63	54
LaNi0.5Cu0.5O3−δ	10	0.55	55
La1−yNi0.6Cu0.4−xFexO3−δ	50	0.55	56
La0.5Sr1.5Ni0.9Cu0.1O4−δ	13.4	0.53	57

---

Single-atom electrocatalysts

Single-atom catalysts, which comprise an isolated active site composed of a single metal atom, have made significant strides in recent years and have been recognized as the most promising electrocatalysts.^58,59 In contrast to conventional nanoparticle catalysts, the utilization of metal sites in an isolated dispersion maximizes atom utilization. This not only results in a substantial cost reduction but also imparts exceptional catalytic activity and selectivity. The study conducted by Akagi et al. examines the potential of anodic catalysts such as Pt, Ru, Ir, Co, Ni, Ta, Fe, and Ti in the electrolysis of liquid NH₃.³³ The current densities were as follows: Ru > Ni > Co > Ir > Pt > Fe at 0.3 V versus H₂/NH₃ (Fig. 7A). The AOR reaction on Ti, Ta, Fe, and Co is constrained by nitrogen desorption, as these elements possess a low metal nitride formation enthalpy (Δ_fH_MN) by their high affinity for nitrogen. However, since Ir and Pt have a greater Δ_fH_MN, H₂ scission or NH₂-adsorption from NH_x to form N atoms restricts the rate at which they occur. At the apex of the volcano correlation, Ru exhibiting a moderate Δ_fH_MN was identified, indicating the greatest catalytic activity³³ (Fig. 7B). Additionally, the authors investigated the impact of Pt and Ni electrode surface configurations, such as plates and films. The current density of the sputtered Pt and Ni films, which possessed a columnar structure and an uneven surface, was observed to be seven and thirteen times greater than that of the Pt and Ni plates, respectively. The sputtered Ni film demonstrated an anodic current density of 0.45 mA cm⁻² at an anodic potential of 0.3 V versus H₂/NH₃ (ref. 33) (Fig. 7C).

Fig. 7 (A) LSV curves at 10 mV s⁻¹, (B) volcano plot of current density at 0.3 V vs. H₂/NH₃ as a function of metal nitride formation enthalpy (Δ_fH_MN (kJ mol⁻¹ N⁻¹)) of various metal plates, and (C) current density for Pt and Ni plates and films for anodic reaction in liquid NH₃ with 0.5 M KNH₂. Reproduced from ref. 33 with permission from Elsevier, copyright 2022.

The Pt catalyst is the most frequently utilized noble metal catalyst in AOR research.²⁴ The selection of an ideal electrolyte is critical in optimizing the functionality of electrocatalysts in AOR. In addition to providing the electrolyte with the ions required for the electrochemical reaction, it also affects the stability of the catalyst, the rate of the reaction, and the selectivity of the product. In general, the majority of researchers employed non-aqueous electrolytes; nevertheless, this approach resulted in a rapid decrease in current due to the electrode becoming contaminated with nitride that formed on the anode surface. The incorporation of methanol into the PF₆–DMF (non-aqueous electrolyte), as suggested by Yang et al., significantly improves the electrochemical characteristics of the system, including NH₃ electrolysis stability reduced polarization resistance, and high current density.⁴¹ The NH₃ electrolysis is conducted for 2 hours in the absence of methanol, during which the current density stabilizes at approximately 0.24 mA cm⁻². The abrupt increase in current density occurs when 5 wt% methanol is added to the KPF₆–DMF non-aqueous electrolyte. The stabilized current density of 1.45 mA cm⁻² is exactly six times that of the KPF₆–DMF non-aqueous electrolyte when methanol is not present. Methanol diminishes the intensity of the adsorption of N* species onto the Pt electrode surface, thereby impeding the formation of a nitridation layer.⁶⁰

In recent times, considerable effort has been devoted to modifying the surface orientation and morphologies of Pt particles in an attempt to expose a greater number of active sites. As an illustration, Liu et al. investigated different morphologies of Pt architectures including flower, sheet, cauliflower-like shapes, and prickly spheres, to demonstrate that the AOR activity is significantly influenced by the Pt surface morphology. Due to better diffusion control, prickly nanospheres had a current density of 2.6 mA cm⁻² and better AOR performance than smooth nanoparticles³⁷ (Fig. 8A). Both the high SSA and the high electrocatalytic activity per unit SSA of smooth, spherical Pt particles explain why they are so much more active as electrocatalysts than hierarchical Pt particles³⁷ (Fig. 8B). Because of the high electrochemically accessible area of the electrocatalyst and the increase of the interaction with water, nitrogen-doped carbon (2.4 mA cm⁻²) and Rh surface entities (2.5 mA cm⁻²) can improve the AOR on Pt nanoparticles.^38,39 An analogous result was observed when PtIrCu hyperbranched concave octahedral nanodendrites improved the electronic effect, exhibiting the highest current density (122.9 A g⁻¹) and the onset potential (0.35 V vs. RHE).⁴⁰ The exceptional performance of PtIrCu can also be ascribed to the presence of profuse suspended bonds on the high-index facets {553}, {331}, and {221}, which readily interact with NH₃ and promote NH₃ oxidation. In a separate study, Yang et al. documented the characteristics of a rhombic dodecahedron nanoframe composed of PtIrCu featuring high-index faceted hyperbranched nanodendrites (RDNF-HNDs). These nanodendrites exhibited a high mass activity of 26.1 A g_PtIr⁻¹ at 0.50 V, which is 7.5 times greater than the mass activity of commercial Pt/C in the AOR 36 and 140 mV lower than RHE.

Fig. 8 (A) Schematic representation of a two-dimensional map illustrating how the voltage difference and scan rate affect the particle morphologies; and (B) Pt particles with varied morphologies and EL values of −0.1 V and −0.6 V have varying electrocatalytic activity per unit SSA. Reproduced from ref. 37 with permission from Elsevier, copyright 2013.

AOR is influenced by the surface orientation of Pt nanocrystals. AOR occurs mostly on the Pt (100) crystalline plane rather than on the Pt (111) and (110) planes.^14,28,61,62 In contrast to low-index surfaces, high-index crystal surfaces are adorned with an abundance of suspended bonds that facilitate reaction with reactants and thereby augment catalytic activity.^63,64 For AOR, Vidal-Iglesias et al. utilized PtMe (Me: Pd, Rh, Ir, Ru) with a particular orientation of Pt (1 0 0) nanoparticles. Out of all the bimetallic samples that were examined, only PtIr and PtRh nanoparticles exhibited an increase in the oxidation density current at low potentials, in contrast to the behavior observed in pure Pt nanoparticles synthesized using the identical technique. PtIr and PtRh nanoparticles, which are preferentially oriented with Pt (100), exhibit greater current densities than polycrystalline Pt nanoparticles. This is because AOR is more sensitive to the presence of square-symmetrical surface sites. On roughened surfaces, however, an increased proportion of exposed Pt(311) planes can substantially increase (by up to 550%) the activity of Pt toward AOR.⁶⁵

Furthermore, one way to manipulate electrical structures is by alloying Pt with other metals.⁶⁶ Pt is alloyed with noble metals like PtPd, PtRh, PtRu, and PtIr to increase the AOR's catalytic activity and lower the onset potential by lowering the dehydrogenation reaction's dissociation energy.^42,67–70 The addition of Ir to Pt, for instance, can substantially decrease the activation energy of NH₃ on Pt, thereby decreasing the AOR overpotential.⁷¹ On the other hand, according to the free energy calculations, the onset potential for AOR is lower on Ir compared to Pt. Ir-coated Pt nanocubes, according to Siddharth et al., are the surface-engineered model catalysts for AOR.⁹ A comparison of adsorption energies revealed that reaction intermediates, specifically *NH₃, bind more strongly to Ir-decorated Pt (100) with a binding energy of −1.68 eV than to Pt (100) with a binding energy of −1.44 eV. This stronger adsorption on Ir-decorated Pt (100) suggests that the catalyst could enhance AOR kinetics. Fig. 9A shows the Gibbs free energy curves for AOR pathways on Pt(100) and Ir-decorated Pt(100). Among the three pathways (N₂H₂, N₂H₃ and N₂H₄), a N₂H₃ pathway is the most favourable one, where the *NH₂-to-*NH is the rate-determining step with the energy barrier of 0.96 eV for Pt (100).⁹ Interestingly, a reduced energy barrier of 0.74 eV for *NH₂-to-*NH step is observed for Ir-decorated Pt(100). They also investigated the optimized structure of AOR intermediates and discovered that there are five optimized structures for favorable N₂H₃ pathways on Ir-decorated Pt(100). Comparing the energy levels in Fig. 9B reveals that a greater binding to the chemical intermediates is facilitated by the surface Ir.⁹ Simulation results show that Ir-decorated Pt NCs/C promote reaction intermediate adsorption and lower the energy barrier during the potential determining phase, increasing their activity.

Fig. 9 (A) Gibbs free energy curves for AOR pathways on Pt(100) and Ir-decorated Pt(100) and (B) optimized structures (top and side views) of reaction intermediates using the N₂H₃ route during AOR on Ir-decorated Pt(100). Color code: Ir, pink; Pt, dark blue; H, white; N, blue. Reproduced from ref. 9 with permission from Springer, copyright 2020.

In addition, adding 3D transition metals like PtIrZn and PtIrNi to PtIr alloy increased AOR activity.^43,72 An instance of this can be seen in the PtIrZn electrode, which exhibits a low onset potential of 0.30 V relative to RHE and a high exchange current density of 0.0043 mA cm⁻². This indicates that in contrast to the nanostructuration of PtIr, the most probable source is the introduction of Zn. H₂ electrochemistry takes place on Pt-based electrodes between 0 and 0.40 V; therefore, dissociative adsorption of NH₃ ought to compete with H₂ adsorption. The inhibition of competitive H₂ adsorption would be advantageous. This may be accomplished by introducing Zn with a high H₂ overpotential into PtIr.

For PtIrNi, the catalyst showed significantly enhanced AOR activity with a reduced onset potential of approximately 0.40 V, compared to the binary PtIr and bare Pt catalysts. This improvement can be explained by the d-band model. The electronic structure analysis revealed that Ir atoms on the surface have a higher d-band center relative to Pt atoms (−1.97 vs. −2.25 eV), resulting in fewer occupied antibonding states and stronger adsorption energies for *NH⁷² (Fig. 10A). Adding Ni to the surface further shifts the d-orbital density of states up in energy (−1.97 to −1.87 eV), strengthening the adsorption of *NH intermediates and enhancing AOR activity. This group orbital model theoretically supports the promotional role of Ni in PtIr alloys for improved AOR performance. The Gibbs free energy pathways also supported that with the addition of a Ni atom to the Pt₃Ir(100) surface, the dehydrogenation of *NH₂ to *NH becomes exergonic at 0.3 V vs. RHE which lowering the thermodynamic barrier in the dehydrogenation process⁷² (Fig. 10B).

Fig. 10 (A) Group d-orbital density of states projected onto the bridge site of model systems and (B) free-energy pathway via constant-potential DFT calculations for initial ammonia dehydrogenation steps at 0.3 V vs. RHE on Pt(100), Pt₃Ir(100), Pt₃IrNi(100). Reproduced from ref. 72 with permission from ACS Publications, copyright 2020.

In addition to promoting alloy formation, Pt-supported metal oxide at the nanoscale can also accelerate the AOR reaction. Ntais et al. accomplished the synthesis of platinum nanoparticles that were backed by NiO, MnO₂, and carbon black.⁴⁴ As compared to Hg/HgO, Pt/NiO exhibited the greatest current density (0.24 mA cm⁻²) and the least onset potential for AOR (−0.48 V). Similarly, Pt supported on TiO₂ nanoparticles and embellished with sulfur-doped onion-like carbon nanoparticles (S-OLCNs) exhibited superior resistance to charge transfer and enhanced long-term stability.⁴⁵ Pt/TiO₂/S-OLCN demonstrated a current density of 0.28 mA cm⁻², which is greater than that of commercial Pt/C. The enhanced catalytic performance of the Pt/TiO₂/S-OLCN electrocatalyst can be attributed to the cohesive metal–support interface formed by the S-OLCN-TiO₂ composite support material and Pt nanoparticles. Additionally, this may be the result of the “spillover” phenomenon, in which intermediates adsorbed on Pt sites migrate to adjacent Ti–OH species to aid in the acceleration of oxidation. TiO helps keep the Pt catalyst from becoming inactivated by N₂ intermediates (Pt–N_ads), which are harmful to the Pt that is created during the AOR. In addition, the Pt/TiO₂/S-OLCN catalyst will be more able to catalyse the creation of AOR because the Pt NPs and the S groups in the onion structure have robust electron transfer connectivity, which will change the surface's electrical state. Lastly, Barbosa et al. reported that PtSnO₂ nanoparticles with a preferential (100) orientation significantly decreased the onset potential for the AOR by approximately 70 mV.¹⁴ The high catalytic activity of these materials is attributed to the Pt (100) orientation, which exhibits a higher tolerance for poisoning intermediates like adsorbed nitrogen species (N_ads). Additionally, the presence of SnO₂ helps prevent the deactivation of the Pt surface caused by intermediates formed during the AOR.

Recently, Zhong et al. conducted a study on a series of transition metal atoms anchored at a g-CN monolayer (TM₂@g-CN), including Fe, Co, Ni, Ru, Rh, Pd, Os, Ir, and Pt, to identify high-performance AOR electrocatalysts.⁴⁶ Compared with bulk Pt, the interaction between metal and support on the TM₂@g-CN catalyst leads to a positive charge on the metal atoms, making their corresponding d orbitals easier to polarize and more effective at adsorbing NH₃ (ref. 46) (Fig. 11A). This interaction also accelerates the coupling of dinitrogen, reducing the onset potential of NH₃ oxidation to 0.48–0.52 eV, compared to the experimentally measured onset potential of 0.62 eV for the Pt (100) surface, without producing surface-poisonous nitrogen. Among all TM₂@g-CN electrocatalysts, Fe, Co, Ru, Rh, and Ir anchored in the g-CN monolayer shows the most promise as AOR catalysts due to their low limiting potentials of −0.47, −0.5, −0.48, −0.52, and −0.48 V, respectively. Furthermore, the catalytic performance of these TM₂@g-CN candidates strongly correlate with the d band center (ε_d) of the TM atoms, which is beneficial for designing promising AOR catalysts. When the ε_d shifts closer to the Fermi energy level, the limiting potential (U_L) value becomes less negative, indicating increased activity⁴⁶ (Fig. 11B). Therefore, the ε_d can directly predict AOR activity, and a downshift in the metal ε_d helps maintain long-lasting electrocatalytic activity.

Fig. 11 (A) Adsorption energy of transition metal atoms anchored at a g-CN monolayer (TM₂@g-CN) and (B) the scaling relationship between d band center (ε_d) and the limiting potential (U_L). Reproduced from ref. 46 with permission from Elsevier, copyright 2023.

Based on the activation energy barrier, the dimerization of NH₂ in Ir₂@g-CN is more favorable than in other metals, with an activation energy of 0.93 eV. The activation energies for the other four TM₂@g-CN catalysts are 2.17, 1.76, 3.3, and 1.59 eV, all higher than the activation energy for the Pt (111) surface, which is 1.07 eV. Referring to the adsorption and activation energy, the mechanism proposed by G–M is kinetically preferred. This mechanism involves N–N bond formation via hydrogenated NH_x species rather than N adatoms because the strong adsorption energy of NH₂ on the TM₂@g-CN monolayer requires a large potential for NH₂ deprotonation into NH. Therefore, the N + N mechanism is not considered in this study.

In summary, recent research focuses on optimizing catalyst materials, morphologies, and surface orientations to enhance AOR performance, with a particular emphasis on single-atom catalysts and alloying strategies.

Transition metal-based electrocatalysts

Active catalysts in AOR have been regarded as Pt and Pt-based noble metals due to their high N₂ selectivity and low overpotential; nevertheless, their substantial expenses and vulnerability to deactivation have demonstrated themselves to be formidable obstacles. As an alternative to platinum-based electrocatalysts, non-noble metal-based electrocatalysts have been suggested recently. Non-noble metal catalysts, particularly transition metals (e.g., Ni-based compounds, CuO, Ni–Co bimetallic and Ni–Cu), are of interest in AOR for electrolytic H₂ production and effluent cleanup.⁷³ After conducting a large number of theoretical calculations and practical investigations, it was shown that Ni possesses the highest HER exchange current density and the lowest free energy of H₂ adsorption (*GH) among a wide variety of nonprecious metals. Based on these findings, it appears that Ni could be a potential candidate for significantly increasing the intrinsic activity of HER. On the other hand, Yao et al. found that pure nickel had a low amount of AOR activity. This was because pure nickel had a high electrode consumption at the voltage that was required for NH₃ oxidation.⁷⁴ It is not possible for NiO to effectively oxidize NH₃ because the negative charge on the cation vacancies causes the generation of reverse electrons, which speeds up the deactivation of the anion.

In a recent study, Almomani et al. established that the addition of TiO₂ to a Ni-based electrode substantially improves its efficacy for AOR.¹⁰ The redox potential of Ni(II)/Ni(III) increased from +0.65 to 1.24 V as a result of the pretreatment technique for NiO nanocatalyst. This is in contrast to the redox potential of HgO/Hg, which remained unchanged. The NH₃ oxidation potential is the component that is responsible for this increase. As an additional point of interest, it was observed that a potential of 1.24 V is sufficient to fully facilitate the transition of Ni(OH)₂ into NiOOH, which is a situation that encourages the oxidation of NH₃. In the interim, NiO–TiO₂-treated nanopowders demonstrated a notable enhancement in their ability to achieve AOR. As the redox potential for the Ni(II)/Ni(III) reaction increased from +0.72 to 1.35 V, a strongly oxidising environment was formed. This resulted in a rise in the built-on Ni(OH)₂ layer, which in turn led to the oxidation of a bigger quantity of NH₃. In fact, metallic Ni and NiOOH are not catalytically active unless they are electrochemically converted to NiOOH at approximately 1.3 V vs. RHE. The dehydrogenation of NH₃ begins with an electron-coupled proton transfer to NiOOH, resulting in a would-be reversible reduction of NiOOH to Ni(OH)₂.⁷⁵ Huang et al. reported that Cu₂O wire-in-Ni(OH)₂ plate passivated by a thin CuO surface (Ni(OH)₂–Cu₂O@CuO) achieves a stable current density as high as ∼80 mA cm⁻² for more than 30 h, demonstrating a much superior AOR performance than the previous works.⁵¹ This novel wire-in-plate nanostructure, featuring intimate contact among Ni(OH)₂, Cu₂O, and CuO, develops dual interfaces of Ni(OH)₂–Cu₂O and Cu₂O@CuO. These interfaces effectively protect Cu₂O from oxidation and inhibit the dissociation of copper species into the electrolyte. In addition, after adding 1 M NH₃ into the electrolyte, a large current density of ∼60 mA cm⁻² was detected with an onset potential of 0.47 V vs. Hg/HgO for 10 mA cm⁻².

Similarly, Xu et al. found that Cu effectively improve activity of Ni(OH)₂ for NH₃ electrooxidation which mixed NiCu layered hydroxides achieve a current density of 35 mA cm⁻² at 0.55 V vs. Ag/AgCl and 0.28 V of onset potential at higher NaOH concentration, which is much higher than that of bare Ni(OH)₂ catalyst (5 mA cm⁻²).⁵² This is attributed to the abundant active sites and the synergistic effect between Ni and Cu, likely due to the formation of Ni_1−xCu_xOOH on the surface of the catalysts through the electrochemical activation of the mixture of Cu(OH)₂ and Ni(OH)₂. Xu et al. also claimed that NiOOH is the actual active substance.⁷⁶ NiOOH is formed from β-Ni(OH)₂via an electrochemical reaction in an alkaline solution to catalyze NH₃ oxidation and then is reduced back to β-Ni(OH)₂ (eqn (15) and (16)). This study showed that without Ni element, the Cu catalyst was found to have negligible activity.

β-Ni(OH)₂ + OH⁻ → NiOOH + H₂O + e⁻ (15)

NiOOH + NH₃ → β-Ni(OH)₂ + products (16)

In details, Tsai et al. reported that Ni foam (NF) exhibited a redox couple of α-Ni(OH)₂(s) ↔ γ-NiOOH(s) at 0.65 V and 0.38 V on the positive and negative scans.⁵³ However, an irreversible anodic peak at approximately 0.95 V (vs. Hg/HgO) was observed when the addition of Cu₂O with different three morphologies of flower, particle, and sheet Cu₂O. This result indicated that the transformation of β-Ni(OH)₂(s) to β-NiOOH(s) via the oxidation process with the addition of Cu₂O. At anodic peaks around 0.95 V, the electrodes demonstrated varying peak current densities, highlighting their catalytic activity. Flower Cu₂O/NF had the highest peak current density at 4.6 mA cm⁻², followed by particle Cu₂O/NF at 3.1 mA cm⁻², sheet Cu₂O/NF at 2.2 mA cm⁻², and NF at 1.2 mA cm⁻². This order reveals that the presence of Cu₂O in combination with NF significantly enhances catalytic activity. In terms of onset potential, it decreased from 0.75 V (for the NF electrode) to 0.55 V for all Cu₂O/NF electrodes. This indicates an increase in the NF electrode's activity towards AOR with the introduction of Cu₂O electrocatalysts.

Co and its composites, which are equally as extensively utilized as Ni, were also the subject of research due to their exceptional H₂ production activity, high electrical conductivity, and durability. After that, the introduction of Co into Ni–C resulted in an improvement in AOR through the reduction of oxidation peaks within the potential range of 0.23–0.56 V, the augmentation of reversible metal-hydroxide production, and the generation of power (P_gen) to maintain the reaction.⁴⁸ Greeley et al. concluded that, in addition to noble metals such as Pt and Ir, transition metals Ni and Co demonstrated high activities for HER.⁷⁷

Metal oxides often suffer from insufficient activity and poor electrical conductivity. However, adding electronegative non-metals like phosphorus (P) and sulfur (S) can modify the electronic structure of transition metals and accelerate the surface reconstruction of the catalyst. The high electronegativity of S attracts electrons from the outer layer of the transition metal. This attraction lowers the energy barrier for reaction intermediates, thereby enhancing the catalytic properties. This approach also significantly enhances the electrical conductivity of the catalysts. For example, a heterostructured electrocatalyst, Ni₃S₄@NiCo₂O₄/NF, was synthesized using the hydrothermal method for the AOR reaction.⁵⁴ Electrochemical tests indicated that the heterostructure of Ni₃S₄ and NiCo₂O₄ enhanced both the electrocatalytic activity and electrical conductivity compared to single-component catalysts, which demonstrated the highest current response and the lowest onset potential of 0.63 V vs. Hg/HgO at the current density of 50 mA cm⁻². In situ Raman spectroscopy revealed that during the AOR process, the surface of the heterostructured catalyst underwent significant reconstruction, forming γ-NiOOH with a more loose and disordered structure. Additionally, the NiCo₂O₄ phase facilitated reconstruction at lower potentials. DFT calculations demonstrated that electron transfer within the heterostructured catalysts accelerated the reconstruction process and reduced the energy barriers for Ni₃S₄ and NiCo₂O₄ in the *NHNH₂ generation step. This promoted NH₃ adsorption and N₂ desorption, resulting in excellent AOR performance and following the G–M mechanism. The production of N₂ in the AOR process represents a rapid energy release mechanism that effectively mitigates critical catalyst poisoning issues.

Recently, engineering confined transition metal nitride nanostructures offers an effective approach to directly control the NH₃ adsorption/desorption process. This is facilitated by their tunable metal composition, excellent conductivity, high corrosion resistance, and large surface area. For example, after conducting electrochemical measurements, it was determined that the 1D nanoneedle Ni–Co₂N electrode exhibited favourable surface area and active sites. This was demonstrated by its near-zero H₂ evolution onset potential, 74 mV overpotential at 10 mA cm⁻², 138 mV at 100 mA cm⁻², and 67 mV dec⁻¹ Tafel slope.²¹ Furthermore, it exhibited exceptional long-term stability, as evidenced by a marginal decrease in the rate of H₂ production following a decade of chronoamperometry measurements. In addition, at 100 mA cm⁻², the potential of NH₃ electrolysis in a two-electrode system was 0.71 V, which was lower than that of water splitting. This indicates that the H₂ produced from NH₃ electrolysis may represent a viable alternative to water splitting. Interestingly, the incorporation of transition metal nitrides results in a downshift of the metal d-band center.⁷⁸ This effectively increases the antibonding electron density in the metal-adsorbate bond, potentially enabling electrocatalysts to achieve targeted moderate metal-adsorbate interactions through d-band contraction. Theoretical simulations indicate that the downshift of the metal d-band on nickel-cobalt nitride (NiCo₂N) nanosheets helps maintain long-lasting AOR activity.¹⁵ This is due to the moderate metal–NH₃ bond strength resulting from the d-band center downshift, optimizing key steps in AOR such as NH₃ adsorption and *N₂ species desorption. Nitriding metals downshift metal d-bands and create empty e_g and t_2g states near the Fermi level. These empty states increase the oxidation states of the metal center and act as electron acceptors, reducing the barrier to generate NH_x⁺ cation radicals by facilitating fast electron transfer from the adsorbed NH_x to the high oxidation state metal site. NiCo₂N nanosheets demonstrate lower charge transfer resistance (4.5 Ω) during AOR and a lower onset overpotential of 0.55 V.

Aside from that, the dendrite-like NiCuBO/NF demonstrates exceptional stability and electrocatalytic activity for AOR. Attainment of 25 mA cm⁻² current density and a Tafel slope of 63 mV dec⁻¹ for AOR requires a mere 1.301 V.⁴⁹ In a similar vein, boron nanoclusters deposited onto NiFe layered double hydroxide nanosheets (B-NiFe-LDH/NF) demonstrated notable efficacy, as evidenced by a current density of 100 mA cm⁻² and a voltage of 1.781 V, which is 202 mV lower than the voltage necessary for complete water splitting. Furthermore, after 24 hours of electrolysis, 78.6% of the NH₃ was removed⁵⁰ (Fig. 12A and B). Furthermore, the utilization of density functional theory (DFT) simulation revealed that the formation of boron nanoclusters resulted in a reduction of the energy barrier, thereby promoting the NH₃ adsorption–desorption process and the Volmer phase.

Fig. 12 (A) Schematic diagram and (B) LSV curve of B-NiFe-LDH/NF electrode. Reproduced from ref. 50 with permission from Elsevier, copyright 2023.

Fig. 13A illustrates that the effective AOR activity of TiO/Ti foil is demonstrated in an alkaline NH₃ solution at a current density of 0.06 mA cm⁻² and a voltage greater than 0.4 V.³² The surface Vo in TiO is shown to significantly facilitate AOR activity through the reduction of the energy barrier in the rate-determining –HNNH₂ formation phase and the promotion of N₂ desorption, as demonstrated by density functional theory (DFT) calculations³² (Fig. 13B). Defective TiO₂ is a stable and cost-effective catalyst for AOR, which is a potential electrode for efficient H₂ production via NH₃ electrolysis and direct NH₃ fuel cells, as demonstrated by these results. Similarly, the Vo interfacial site is the main adsorption site for NH₃.⁷⁹ Aside from that, Khan et al. reported that the half-filled d-orbitals of Pd and Ru promoted ZnO/Al₂O₃ ternary materials as effective and robust electrocatalysts towards AOR.⁴⁷ The current density was 0.307 mA cm⁻² for PdO/ZnO/Al₂O₃, which was less than that of MnO₂/ZnO/Al₂O₃.

Fig. 13 (A) LSV curves of TiO/Ti foil in 0.5 M NaOH and 0.5 M NaOH with 0.1 M NH₃ solution at a scan rate of 10 mV s⁻¹ and (B) schematic diagram of AOR reaction in the presence of an oxygen vacancy on TiO surface. Reproduced from ref. 32 with permission from Elsevier, copyright 2021.

Perovskite oxides, represented by the formula ABO₃, are intriguing electrocatalysts due to their significant chemical and structural flexibility. In an ideal cubic system, the larger A-site is typically occupied by an alkaline-earth metal with 12-fold oxygen coordination, while the smaller B-site is occupied by a transition metal with 6-fold BO₆ oxygen coordination. Zhang et al. investigated the use of Ni/Cu in the B-site of a stoichiometric LaNi_0.5Cu_0.5O_3−δ perovskite as an AOR catalyst.⁵⁵ Their findings revealed that annealing the material in an argon atmosphere introduced Vo, which significantly enhanced its catalytic activity (onset potential of 0.55 V vs. Ag/AgCl at 10 mA cm⁻²). Similarly, Jeerh et al. investigated the optimal amount of Fe doping in La_1−yNi_0.6Cu_0.4−xFe_xO_3−δ and discovered that La_0.9Ni_0.6Cu_0.35Fe_0.05O_3−δ exhibited superior activity for the AOR.⁵⁶ This was evidenced by a lower onset potential of 0.55 V vs. Ag/AgCl at 50 mA cm⁻², attributed to the presence of Vo introduced by an A-site deficiency, which led to greater exposure of active Ni²⁺ on the surface. However, compared to the Zhang and co-workers’ study, La_0.5Sr_1.5Ni_0.9Cu_0.1O_4−δ can act as a robust AOR anode and achieve 13.4 mA cm⁻² at a lower onset potential of 0.53 V versus Ag/AgCl.⁵⁷ The selective substitution of La with Sr enhances electronic conductivity and increases Vo concentration. Additionally, doping Cu into the B-site of the perovskite fosters a synergistic interaction between nickel and copper, leading to excellent AOR activity.

Fig. 14 shows the comparison of onset potential for various transition metal electrocatalysts for AOR. Among all catalysts, RuO₂–ZnO/Al₂O₃ showed the lowest onset potential, followed by the Ni_1−xCu_xOOH. The half-filled d-orbitals of Ru promoted ZnO/Al₂O₃ ternary materials as effective and robust electrocatalysts towards AOR, meanwhile the abundant active sites and the synergistic effect between Ni and Cu, likely due to the formation of Ni_1−xCu_xOOH on the surface of the catalysts through the electrochemical activation of the mixture of Cu(OH)₂ and Ni(OH)₂. The perovskite oxide also shows a lower onset potential, which is likely due to the presence of Vo, which led to greater exposure of active Ni²⁺ on the surface. Most Ni-based Cu and Co show comparable onset potential for AOR reaction and the larger onset potential is B-NiFe-LDH/NF.

Fig. 14 Comparison of onset potential of various transition metal electrocatalysts under different current densities for AOR.

As a conclusion, both single-atom electrocatalysts and transition metal-based electrocatalysts have shown promising performance in NH₃ electrooxidation. However, single-atom electrocatalysts offer several advantages over traditional transition metal-based catalysts, including higher catalytic activity, selectivity, and stability. The unique electronic and geometric structures of single-atom catalysts provide highly active and stable catalytic sites, leading to enhanced NH₃ conversion efficiency and H₂ production selectivity. Additionally, single-atom electrocatalysts can minimize the use of expensive transition metals, reducing production costs and improving the scalability of NH₃ electrooxidation technology.

Techno-economic analysis for NH₃ electrolysis

It is a promising alternative to use NH₃ for the long-distance storage and transportation of H₂ that is produced from renewable sources. An evaluation of the technological and economic aspects was carried out to ascertain the levelized cost of H₂ (LCOH) that is produced from renewable NH₃ feedstocks as tabulated in Table 2. Lee et al. conducted a study in which they evaluated the costs that are connected with the thermal breakdown of NH₃ to produce H₂ at two distinct rates which are 0.9 m³ h⁻¹ (0.08 kg h⁻¹) for a laboratory scale and 30 m³ h⁻¹ (2.67 kg h⁻¹) for a small-scale H₂ refueling station.⁸⁰ They established a model of the process in their paper and verified its accuracy by comparing it to experimental findings reported by Chung et al.⁸⁶ The cost of H₂ production was assessed by the authors through the utilization of various economic methodologies, including sensitivity analysis, profitability analysis, and itemized cost estimation. By utilising itemised cost estimating and beginning with a green NH₃ pricing of 548 USD per tNH₃, they were able to calculate that the LCOH on the laboratory scale was 9.97 USD per kgH₂, while the LCOH on the refuelling station scale was 6.27 USD per kgH₂.⁸⁰ These figures correspond to reconversion costs of 6.75 USD per kgH₂ and 3.14 USD per kgH₂, respectively, when the cost of NH₃ feedstock is excluded from the LCOH calculation.

Table 2 Techno-economic for H₂ production via AOR

Method	Cost of H2 production (USD per kgH2)	Ref.
Thermal decomposition	9.97	80
Thermal decomposition	5.5	81
Membrane purification	5.98	82
Fired NH3 cracker	5.2	83
Thermal decomposition	2.61	84
Thermal decomposition	2.2	85
Membrane technology	7.72	19
Electrolysis	9.59
Fixed bed technology	8.69

---

Nasharuddin et al. conducted a separate study in which they evaluated the technological and economic feasibility of NH₃ dissociation into H₂ for centralised installations (1000 kgH₂ d⁻¹) and replenishment stations (500 kgH₂ d⁻¹).⁸¹ Due to differences in CAPEX and OPEX for both scales, the authors concluded that breaking NH₃ is economically viable for centralised H₂ production for 5.5 USD per kgH₂.⁸¹ On the other hand, the refuelling station is economically viable for 101 USD per kgH₂. The price of NH₃ was 520 USD per tNH₃, which the authors used as the basis for their study. Furthermore, Lin et al. devised a purification system to generate H₂ through thermal decomposition at an on-site H₂ refueling station supplied with NH₃ is 26.7 kg h⁻¹.⁸² Numerous purification configurations, including PSA, membrane, PSA-membrane, and others, were contrasted in terms of H₂ production costs by the authors. Beginning with an NH₃ cost of 540 USD per tNH₃, the authors came to the conclusion that the PSA followed by the membrane purification system produces the most cost-effective H₂ production cost of 5.98 USD per kgH₂.⁸² This is equivalent to a reconversion cost of 0.91 USD per kgH₂, which is the most affordable method of producing methane.

Furthermore, for large-scale H₂ production, comprehensive modeling was carried out by Makhloufi and colleagues, who determined that the LCOH was 5.2 USD per kg.⁸³ This was determined by assuming that the H₂ output rate was 7560 kg h⁻¹ and that the price of NH₃ was 488 USD per tNH₃.⁸³ A techno-economic modeling study was undertaken by Lullo et al. to examine 32 H₂ transport systems, which comprised five fluids (hythane, gaseous H₂, NH₃, liquid H₂ and LOHC) and three modes of transportation (vehicle pipeline, and rail).⁸⁴ Following the simulation of the process, they concluded that the amount of thermal energy consumed was 4.38 GJ NG per tH₂, while the amount of electrical energy consumed was 81 kW h per tH₂. In their research, the cost of converting H₂ into H₂ (reconversion) was calculated to be 2.61 USD per kgH₂.⁸⁴ Cesaro et al. assessed the expenses associated with generating electricity from NH₃via various methods, including the conversion of NH₃ to H₂ for fuel cell applications.⁸⁷ To determine the energy efficiency and costs associated with the various routes of NH₃ thermal breakdown using a large-scale NH₃ refinery, a techno-economic assessment was carried out. After everything was said and done, they concluded that the levelized cost of energy would be anywhere between 167 and 197 USD per MW h at a plant capacity factor of 25%, presuming that NH₃ would cost 380 USD per tNH₃.

Papadias et al. conducted an assessment of the expenses associated with utilizing various liquid H₂ carriers, such as toluene, methanol, and NH₃, to transport H₂.⁸⁵ At a rate of 50 tpd, the authors found that NH₃ had the highest production cost (2.2 USD per kgH₂) and the lowest breakdown cost to H₂ (0.61 USD per kgH₂).⁸⁵ According to the IEA's “The Future of H₂” report, the anticipated cost of converting NH₃ to H₂ in a centralised facility (15 000 tNH₃ per year) would be 1 USD per kgH₂. However, the report does not provide any specific assumptions regarding the reconversion process.⁸⁸

In a recent study conducted by Kanaan et al., the most economical ways to produce H₂ gas from NH₃ were found to be electrolysis while in transit, thermal decomposition using membrane technology at the import terminal, and a fixed bed at the H₂ refuelling station.⁸⁹ Given the significance of the capital cost in conducting the techno-economic assessment, it is imperative to determine the appropriate dimensions for each piece of equipment. They discovered that membrane reactor technology is currently the most cost-effective method for converting NH₃ to H₂ (7.72 USD per kgH₂). The technology exhibits benefit in comparison to traditional fixed bed technologies by its reduced operational expenses. Electrolysis of NH₃ on a truck resulted in the lowest LCOH, at 9.59 USD per kgH₂. Membrane technology has a somewhat higher LCOH (9.14 USD per kgH₂) than fixed bed technology (8.69 USD per kgH₂).¹⁹ The variations in the process configurations for the various scales account for this. The cost of H₂ gas produced at the import terminal using fixed bed and membrane technology is competitive in terms of LCOH, especially when compared to the current green H₂ pricing in Europe, which are 4–8 USD per kgH₂, according to the IEA.

Summary, future perspectives, and recommendations

Electrocatalysts play a crucial role in the electrooxidation of NH₃ for clean H₂ production. With the increasing demand for renewable energy sources, electrocatalysis has emerged as a promising method for efficient and sustainable H₂ production. NH₃, with its high H₂ content, has gained attention as a potential carrier for H₂. However, the electrooxidation of NH₃ faces challenges such as low conversion efficiency, poor selectivity, and catalyst degradation. The commercialization of AOR relies on the development of robust and efficient electrocatalysts. Because of their high NH₃ affinity and fast AOR kinetics, Pt electrocatalysts have been investigated in depth using a variety of techniques, including alloying, surface morphology modification, and electrocatalytic reaction optimization. In particular, by successfully lowering the energy barrier, a Pt–Ir greatly enhanced the reaction kinetics of AOR.

There has been a lot of talk about using cheaper, non-noble metal catalysts, especially catalysts based on nickel, as a substitute for Pt. When the affinity of NH₃ and intermediates on Ni and Cu was changed, the Ni–Cu bimetal catalysts showed the most promising activity in terms of reaction rate and stability. Nevertheless, the development of potent new catalysts and their continued search for are essential for the practical implementation of AOR. Particularly to develop non-noble metal electrocatalysts for AOR that are economical, several obstacles must be surmounted. Initially, selective AOR in regions with minimal overpotential is critical. Alternatively, AOR competes with competing reactions such as water oxidation, NH₃ oxidation to NO₃⁻ or NO₂, and so on, while traditional electrocatalysts based on non-noble metals showed AOR performance at high overpotentials. Conversely, it was discovered that the introduction of heteroatoms into alloys can modify the electronic configuration of catalysts, leading to a decrease in the energy barrier for NH₃ adsorption and eventually a reduction in the overpotential.

Improving AOR requires not just creating catalysts, but also a basic understanding of the reaction mechanism. At present, the AOR mechanism is primarily being studied in relation to its applicability to Pt catalysts. Nevertheless, the specific rate-determining steps or reaction mechanisms may vary among distinct electrocatalyst types. The DFT calculations have been extensively utilized in order to delineate the fundamental effect of the catalytic modification and to propose reaction mechanisms. In light of developments in theoretical inquiry, experimental observation of reaction mechanisms through the monitoring of adsorbed intermediates has become crucial. Recent studies have documented endeavors to examine the true mechanisms of action on electrocatalysts through the utilization of in situ or operando characterizations. In light of the advancements made thus far, the AOR possesses considerable practical development potential through the integration of device engineering, catalytic modification, and fundamental research.

The replacement of water with liquid NH₃ is another feature that could revolutionize AOR technology for fuel cells. NH₃ only becomes corrosive when reacting with water to generate the caustic OH⁻, however, this gas has a far larger energy density and is resistant to pipeline corrosion. The challenges associated with water oxidation are eliminated when a nonaqueous electrolyte is combined with liquid NH₃. Furthermore, it is critical to conduct further mechanistic investigations on nonaqueous liquid NH₃, given its considerable potential provided that the challenges impeding its application in AOR can be resolved. Aside from this, the efficient energy generation via direct NH₃ fuel cells and the electrochemical synthesis of “green NH₃” presents a highly promising pathway towards carbon neutrality by 2050.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request. Please note that due to confidentiality agreements, certain restrictions apply to the availability of the data.

Conflicts of interest

The authors stated that there are no known conflicts of interest in this article.

Acknowledgements

The authors are gratefully acknowledging to the Universiti Teknologi Malaysia for the Fundamental Research Grant (No. 22H51) and Professional Development Research University Grant (No. 06E92 and PY/2024/00993).

References

1. J. A. Okolie, B. R. Patra, A. Mukherjee, S. Nanda, A. K. Dalai and J. A. Kozinski, Int. J. Hydrogen Energy, 2021, 46, 8885–8905.
2. A. Kovač, M. Paranos and D. Marciuš, Int. J. Hydrogen Energy, 2021, 46, 10016–10035.
3. S. Anwar, F. Khan, Y. Zhang and A. Djire, Int. J. Hydrogen Energy, 2021, 46, 32284–32317.
4. L. Cao, I. K. M. Yu, X. Xiong, D. C. W. Tsang, S. Zhang, J. H. Clark, C. Hu, Y. H. Ng, J. Shang and Y. S. Ok, Environ. Res., 2020, 186, 109547.
5. G. Franchi, M. Capocelli, M. De Falco, V. Piemonte and D. Barba, Membranes, 2020, 10, 10.
6. S. A. Lee, M. G. Lee and H. W. Jang, Sci. China Mater., 2022, 65, 3334–3352.
7. J. Łuczak and M. Lieder, Adv. Colloid Interface Sci., 2023, 319, 102963.
8. Y. Tian, Z. Mao, L. Wang and J. Liang, Small Struct., 2023, 4, 2200266.
9. K. Siddharth, Y. Hong, X. Qin, H. J. Lee, Y. T. Chan, S. Zhu, G. Chen, S.-I. Choi and M. Shao, Appl. Catal., B, 2020, 269, 118821.
10. F. Almomani, R. Bhosale, M. Khraisheh, A. Kumar and M. Tawalbeh, Int. J. Hydrogen Energy, 2020, 45, 10398–10408.
11. R. Wang, H. Liu, K. Zhang, G. Zhang, H. Lan and J. Qu, Chem. Eng. J., 2021, 404, 126795.
12. M.-H. Tsai, T.-C. Chen, Y. Juang, L.-C. Hua and C. Huang, Electrochem. Commun., 2020, 121, 106875.
13. F. Quan, G. Zhan, B. Zhou, C. Ling, X. Wang, W. Shen, J. Li, F. Jia and L. Zhang, J. Environ. Sci., 2023, 125, 544–552.
14. J. R. Barbosa, M. N. Leon, C. M. Fernandes, R. M. Antoniassi, O. C. Alves, E. A. Ponzio and J. C. M. Silva, Appl. Catal., B, 2020, 264, 118458.
15. S. He, Y. Chen, M. Wang, H. Nuomin, P. Novello, X. Li, S. Zhu and J. Liu, Nano Energy, 2021, 80, 105528.
16. Z.-H. Lyu, J. Fu, T. Tang, J. Zhang and J.-S. Hu, EnergyChem, 2023, 5, 100093.
17. J. H. Jang, S. Y. Park, D. H. Youn and Y. J. Jang, Catalysts, 2023, 13, 803.
18. P. Peng, J. Su and H. Breunig, Energy Convers. Manage., 2023, 288, 117166.
19. R. Kanaan, P. H. Affonso Nóbrega, P. Achard and C. Beauger, Renewable Sustainable Energy Rev., 2023, 188, 113784.
20. H. Mashhadimoslem, M. Safarzadeh Khosrowshahi, M. Delpisheh, C. Convery, M. Rezakazemi, T. M. Aminabhavi, M. Kamkar and A. Elkamel, Chem. Eng. J., 2023, 474, 145661.
21. K. Jiang, K. Li, Y.-Q. Liu, S. Lin, Z. Wang, D. Wang and Y. Ye, Electrochim. Acta, 2022, 403, 139700.
22. N. Hassan, A. Jalil, N. Khusnun, A. Ahmad, T. Abdullah, R. Kasmani, N. Norazahar, M. Kamaroddin and D. Vo, Environ. Chem. Lett., 2022, 1–23.
23. M. Sawal, A. Jalil, T. Abdullah, N. Khusnun, N. Hassan, F. Aziz, A. Fauzi, M. Kamaroddin, M. Omar and S. Haron, Energy Convers. Manage., 2022, 274, 116456.
24. X. Yang, H. Sun, C. Liu, L. Yu and H. Chen, Chem. Eng. J., 2022, 442, 136167.
25. N. M. Adli, H. Zhang, S. Mukherjee and G. Wu, J. Electrochem. Soc., 2018, 165, J3130–J3147.
26. Y. Tian, H. Tan, X. Li, J. Jia, Z. Mao, J. Liu and J. Liang, Chin. J. Catal., 2024, 56, 25–50.
27. H. Kim, S. Hong, H. Kim, Y. Jun, S. Y. Kim and S. H. Ahn, Appl. Mater. Today, 2022, 29, 101640.
28. K. Siddharth, Y. Chan, L. Wang and M. Shao, Curr. Opin. Electrochem., 2018, 9, 151–157.
29. X. Jiang, D. Ying, X. Liu, M. Liu, S. Zhou, C. Guo, G. Zhao, Y. Wang and J. Jia, Electrochim. Acta, 2020, 345, 136157.
30. J. Huang, Z. Chen, J. Cai, Y. Jin, T. Wang and J. Wang, Nano Res., 2022, 15, 5987–5994.
31. X. Yang, L. Sun, X. Liu, Z. Yang, H. Sun, W. Liu and H. Chen, ACS Catal., 2024, 14, 6236–6246.
32. S. Zhang, Y. Zhao, L. Yan, H. Jiang, X. Yang, Y. Wang, H. Song and X. Zhao, Int. J. Hydrogen Energy, 2021, 46, 39208–39215.
33. N. Akagi, K. Hori, H. Sugime, S. Noda and N. Hanada, J. Catal., 2022, 406, 222–230.
34. D. T. Tran, T. H. Nguyen, H. Jeong, P. K. L. Tran, D. Malhotra, K. U. Jeong, N. H. Kim and J. H. Lee, Nano Energy, 2022, 94, 106929.
35. C. Jo, S. Surendran, M.-C. Kim, T.-Y. An, Y. Lim, H. Choi, G. Janani, S. Cyril Jesudass, D. Jun Moon, J. Kim, J. Young Kim, C. Hyuck Choi, M. Kim, J. Kyu Kim and U. Sim, Chem. Eng. J., 2023, 463, 142314.
36. J. H. Jang, S. Y. Park, D. H. Youn and Y. J. Jang, Catalysts, 2023, 13, 803.
37. J. Liu, W. Hu, C. Zhong and Y. F. Cheng, J. Power Sources, 2013, 223, 165–174.
38. N. N. Fomena, S. Garbarino, E. Bertin, A. Korinek, G. Botton, L. Roué and D. Guay, J. Catal., 2017, 354, 270–277.
39. V. A. Ribeiro, I. C. de Freitas, A. O. Neto, E. V. Spinacé and J. C. M. Silva, Mater. Chem. Phys., 2017, 200, 354–360.
40. X. Lin, X. Zhang, Z. Wang, X. Zhu, J. Zhu, P. Chen, T. Lyu, C. Li, Z. Q. Tian and P. K. Shen, J. Colloid Interface Sci., 2021, 601, 1–11.
41. H. Yang, X. Huang, Z. Liu, X. Lin, Q. Chen, J. Li, C. Zhang, Z. P. Kan, Z. Q. Tian and P. K. Shen, J. Colloid Interface Sci., 2023, 652, 1764–1774.
42. J. C. M. Silva, S. Ntais, É. Teixeira-Neto, E. V. Spinacé, X. Cui, A. O. Neto and E. A. Baranova, Int. J. Hydrogen Energy, 2017, 42, 193–201.
43. J. Jiang, Electrochem. Commun., 2017, 75, 52–55.
44. S. Ntais, A. Serov, N. I. Andersen, A. J. Roy, E. Cossar, A. Allagui, Z. Lu, X. Cui, E. A. Baranova and P. Atanassov, Electrochim. Acta, 2016, 222, 1455–1463.
45. L. L. Sikeyi, L. R. Bila, T. D. Ntuli, C. T. Selepe, N. W. Maxakato, N. J. Coville and M. S. Maubane-Nkadimeng, Diamond Relat. Mater., 2023, 132, 109612.
46. J.-J. Zhong, S.-P. Huang, J.-F. Gu, Y. Li, K.-N. Ding, Y.-F. Zhang, W. Lin and W.-K. Chen, Appl. Surf. Sci., 2023, 609, 155280.
47. S. Khan, S. S. Shah, A. B. Yurtcan, A. A. A. Bahajjaj, A. Zafar and N. K. Janjua, Fuel, 2023, 347, 128446.
48. F. Almomani and M. A. H. S. Saad, Int. J. Hydrogen Energy, 2021, 46, 4678–4690.
49. A. A. Kashale, C.-T. Wu, H.-F. Hsu and I.-W. P. Chen, Chem. Eng. J., 2023, 474, 145907.
50. J. Wang, S. Qing, X. Tong, K. Zhang, G. Luo, J. Ding and L. Xu, Appl. Surf. Sci., 2023, 640, 158330.
51. J. Huang, J. Cai and J. Wang, ACS Appl. Energy Mater., 2020, 3, 4108–4113.
52. W. Xu, R. Lan, D. Du, J. Humphreys, M. Walker, Z. Wu, H. Wang and S. Tao, Appl. Catal., B, 2017, 218, 470–479.
53. M.-H. Tsai, Y. Juang, C.-C. Hu, L.-C. Hua and C. Huang, J. Environ. Chem. Eng., 2024, 12, 112339.
54. L. Wang, K. Jiang, Z. Wang, T. Li, D. Wang and Y.-Q. Liu, Chem. Eng. J., 2024, 492, 152268.
55. M. Zhang, H. Li, X. Duan, P. Zou, G. Jeerh, B. Sun, S. Chen, J. Humphreys, M. Walker and K. Xie, Adv. Sci., 2021, 8, 2101299.
56. G. Jeerh, P. Zou, M. Zhang, S. Chen, J. Humphreys and S. Tao, Sep. Purif. Technol., 2022, 297, 121451.
57. M. Zhang, P. Zou, G. Jeerh, B. Sun, M. Walker and S. Tao, Adv. Funct. Mater., 2022, 32, 2204881.
58. H. Li, W. Wang, S. Xue, J. He, C. Liu, G. Gao, S. Di, S. Wang, J. Wang and Z. Yu, J. Am. Chem. Soc., 2024, 146, 9124–9133.
59. H. Li, S. Di, P. Niu, S. Wang, J. Wang and L. Li, Energy Environ. Sci., 2022, 15, 1601–1610.
60. C. Liu, F. Yang, A. Schechter and L. Feng, Adv. Sens. Energy Mater., 2023, 100055.
61. F. J. Vidal-Iglesias, J. Solla-Gullón, V. Montiel, J. Feliu and A. Aldaz, J. Power Sources, 2007, 171, 448–456.
62. H. S. Pillai and H. Xin, Ind. Eng. Chem. Res., 2019, 58, 10819–10828.
63. C. Xiao, B.-A. Lu, P. Xue, N. Tian, Z.-Y. Zhou, X. Lin, W.-F. Lin and S.-G. Sun, Joule, 2020, 4, 2562–2598.
64. F. J. Vidal-Iglesias, J. Solla-Gullón, P. Rodrıguez, E. Herrero, V. Montiel, J. Feliu and A. Aldaz, Electrochem. Commun., 2004, 6, 1080–1084.
65. S. Johnston, B. H. Suryanto and D. R. MacFarlane, Electrochim. Acta, 2019, 297, 778–783.
66. J. Zhou, J. S. Chung and S. G. Kang, Int. J. Hydrogen Energy, 2024, 58, 745–752.
67. M. H. Assumpção, S. G. da Silva, R. F. de Souza, G. S. Buzzo, E. V. Spinacé, A. O. Neto and J. C. M. Silva, Int. J. Hydrogen Energy, 2014, 39, 5148–5152.
68. M. H. Assumpção, R. M. Piasentin, P. Hammer, R. F. De Souza, G. S. Buzzo, M. C. Santos, E. V. Spinacé, A. O. Neto and J. C. M. Silva, Appl. Catal., B, 2015, 174, 136–144.
69. T. L. Lomocso and E. A. Baranova, Electrochim. Acta, 2011, 56, 8551–8558.
70. Q. Xue, Y. Zhao, J. Zhu, Y. Ding, T. Wang, H. Sun, F. Li, P. Chen, P. Jin and S. Yin, J. Mater. Chem. A, 2021, 9, 8444–8451.
71. A. Estejab and G. G. Botte, Mol. Catal., 2018, 445, 279–292.
72. Y. Li, X. Li, H. S. Pillai, J. Lattimer, N. Mohd Adli, S. Karakalos, M. Chen, L. Guo, H. Xu and J. Yang, ACS Catal., 2020, 10, 3945–3957.
73. A. Allagui, S. Sarfraz and E. A. Baranova, Electrochim. Acta, 2013, 110, 253–259.
74. K. Yao and Y. Cheng, Mater. Chem. Phys., 2008, 108, 247–250.
75. J. Łuczak and M. Lieder, Adv. Colloid Interface Sci., 2023, 102963.
76. W. Xu, D. Du, R. Lan, J. Humphreys, D. N. Miller, M. Walker, Z. Wu, J. T. Irvine and S. Tao, Appl. Catal., B, 2018, 237, 1101–1109.
77. J. Greeley, T. F. Jaramillo, J. Bonde, I. Chorkendorff and J. K. Nørskov, Nat. Mater., 2006, 5, 909–913.
78. D. J. Little, D. O. Edwards, M. R. Smith III and T. W. Hamann, ACS Appl. Mater. Interfaces, 2017, 9, 16228–16235.
79. H. Sun, H. Wang and Z. Qu, ACS Catal., 2023, 13, 1077–1088.
80. B. Lee, J. Park, H. Lee, M. Byun, C. W. Yoon and H. Lim, Renewable Sustainable Energy Rev., 2019, 113, 109262.
81. R. Nasharuddin, M. Zhu, Z. Zhang and D. Zhang, Int. J. Hydrogen Energy, 2019, 44, 14445–14455.
82. L. Lin, Y. Tian, W. Su, Y. Luo, C. Chen and L. Jiang, Sustainable Energy Fuels, 2020, 4, 3006–3017.
83. C. Makhloufi and N. Kezibri, Int. J. Hydrogen Energy, 2021, 46, 34777–34787.
84. G. Di Lullo, T. Giwa, A. Okunlola, M. Davis, T. Mehedi, A. Oni and A. Kumar, Int. J. Hydrogen Energy, 2022, 47, 35293–35319.
85. D. D. Papadias, J.-K. Peng and R. K. Ahluwalia, Int. J. Hydrogen Energy, 2021, 46, 24169–24189.
86. D. B. Chung, H. Y. Kim, M. Jeon, D. H. Lee, H. S. Park, S. H. Choi, S. W. Nam, S. C. Jang, J.-H. Park and K.-Y. Lee, Int. J. Hydrogen Energy, 2017, 42, 1639–1647.
87. Z. Cesaro, M. Ives, R. Nayak-Luke, M. Mason and R. Bañares-Alcántara, Appl. Energy, 2021, 282, 116009.
88. IEA, The Future Hydrogen, 2019,  DOI:10.1787/1e0514c4-en.
89. R. Kanaan, P. H. A. Nóbrega, P. Achard and C. Beauger, Renewable Sustainable Energy Rev., 2023, 188, 113784.

---