Design of earth‐abundant amorphous transition metal‐based catalysts for electrooxidation of small molecules: Advances and perspectives

Electrochemical oxidation of small molecules (e.g., water, urea, methanol, hydrazine, and glycerol) has gained growing scientific interest in the fields of electrochemical energy conversion/storage and environmental remediation. Designing cost‐effective catalysts for the electrooxidation of small molecules (ESM) is thus crucial for improving reaction efficiency. Recently, earth‐abundant amorphous transition metal (TM)‐based nanomaterials have aroused souring interest owing to their earth‐abundance, flexible structures, and excellent electrochemical activities. Hundreds of amorphous TM‐based nanomaterials have been designed and used as promising ESM catalysts. Herein, recent advances in the design of amorphous TM‐based ESM catalysts are comprehensively reviewed. The features (e.g., large specific surface area, flexible electronic structure, and facile structure reconstruction) of amorphous TM‐based ESM catalysts are first analyzed. Afterward, the design of various TM‐based catalysts with advanced strategies (e.g., nanostructure design, component regulation, heteroatom doping, and heterostructure construction) is fully scrutinized, and the catalysts’ structure‐performance correlation is emphasized. Future perspectives in the development of cost‐effective amorphous TM‐based catalysts are then outlined. This review is expected to provide practical strategies for the design of next‐generation amorphous electrocatalysts.

and environmental remediation techniques. For example, water oxidation, namely, the oxygen evolution reaction (OER), is a key reaction in water electrolysis-driven hydrogen production techniques and metal-air batteries. 2 The urea/hydrazine oxidation reaction (UOR/HzOR) is important for urine/hydrazine wastewater purification and direct urea/hydrazine fuel cells. 3 Moreover, coupling UOR/HzOR with hydrogen evolution reaction (HER) is suggested to realize energy-saving hydrogen production. 4 Similarly, the electrochemical oxidation of methanol, glycerol, sulfion, formic acid, and 5-hydroxymethylfurfural (HMF) also has been investigated to replace OER for producing hydrogen with lower energy consumption and obtaining value-added chemicals. 5,6 Overall, the oxidative transformation of these small molecules in an electrochemical way would significantly benefit the sustainable energy system and green environment.
ESM generally shows a multi-electron reaction characteristic and leads to high-energy barriers. Therefore, it is important to boost the ESM process with efficient electrocatalysts. Precious metals (e.g., Ru, Ir, Pd, Rh, and Pt)-based catalysts have shown high catalytic activities and good durability, 7,8 but their high prices and low reserves severely hamper their applications. 9,10 Alternatively, catalysts based on abundant transition metals (TMs, e.g., Fe, Ni, and Co) have shown promising catalytic performance for ESM, 11 especially in alkaline or neutral electrolytes. Great efforts have thus been made to exploit cost-efficient TM-based ESM catalysts.
Thousands of TM-based catalysts with diverse nanostructures, crystal structures, and compositions have been developed. Among them, catalysts with an amorphous phase have attracted enormous scientific attention due to their flexible crystal/electronic structures, rich electroactive sites, and so on. Thus, amorphous catalysts show better electrochemical performance for ESM than crystalline counterparts. 12 To upgrade the catalytic performance of amorphous catalysts, efficient strategies have been adopted to populate the catalytically active sites and/or improve the intrinsic catalytic activity of catalysts, such as nanostructure design, component regulation, heteroatom doping, and heterostructure construction. For example, the catalytic activity of copper hydroxide toward HMF oxidation can be significantly enhanced by forming a heterostructure with nickel hydroxide. 13 This is because the presence of nickel hydroxide can enrich the active sites for HMF oxidation. By utilizing these powerful strategies, many efficient catalysts based on amorphous TM alloys, (hydr)oxides, chalcogenides, phosphides, nitrides, borides, and oxometallates have been developed for ESM (Scheme 1 and Table 1). These amorphous materials significantly enrich the ESM catalyst community and largely promote related reaction processes. In consequence, the great interest in this topic hastens a comprehensive summary covering the design of amorphous TM-based ESM electrocatalysts.
This review aims to deliver a critical examination of recent advancements in the design of amorphous

FEATURES OF AMORPHOUS ELECTROCATALYSTS
Amorphous catalysts possess unique characteristics of short-range order and long-range disorder and thereby manifest different physicochemical properties from their crystalline analogs. 50,51 Generally, the better catalytic performance of amorphous catalysts than their crystalline counterparts can be explained by their following features, namely, large specific surface area (SSA), flexible electronic structure, and deep structure reconstruction during the ESM process.

Large specific surface area
Compared with crystalline materials that generally synthesized at relatively high temperatures, amorphous materials prepared at mild conditions generally show higher SSA. 52,53 A main reason is that the amorphousto-crystalline conversion process would lead to particle coarsening during the high-temperature process. 54 This feature of amorphous catalysts guarantees more electroactive sites and a larger electrochemical active surface area (ECSA) and promotes the interaction between catalysts and reactants/intermediates. 50,55 As reported by Cai et al., the amorphous NiFe alloy synthesized at room temperature comprises interconnected nanoparticles with an average size of 50 nm ( Figure 1A,B), which is smaller than the crystalline counterpart prepared at 600 • C (100 nm, Figure 1C,D). 56 The amorphous catalysts with smaller size exhibit higher SSA, which can promote mass/charge transfer and provides abundant transportation channels for electrolytes and gas products. 57 The high SSA of amorphous TM-based catalysts also results in larger ECSA ( Figure 1E), and smaller Tafel slopes ( Figure 1F), thereby faster reaction kinetics. 56 As a large SSA ensures better contact at the catalyst-electrolyte interface, amorphous catalysts show smaller charge transfer resistance (R ct ) than crystalline materials.

Flexible electronic structure
The long-range disorder feature gives rise to flexible electronic structures in amorphous TM-based catalysts, which results in two major advantages. Amorphous catalysts possess more defects or vacancies that are generally electroactive sites than long-range ordered crystalline counterparts. 58,59 As depicted in Figure 2A, the LaNiO 3 catalysts prepared at lower temperatures (450 and 500 • C) possess a hybrid amorphous/crystalline feature, and the amorphous part shows no atomic periodicity. 60 This also leads to rich electronic properties, which can be identified by X-ray photoelectron spectra analysis. Compared with the crystalline NiO x /NF-600 • C catalyst, the Ni 2p3/2 spectrum of amorphous NiO x /NF-100 • C contains a higher ratio of high-valence Ni 3+ species. In addition, the amorphous NiO x /NF-100 • C exhibits a higher concentration of oxygen defects. These favorable electronic features benefit the adsorption/desorption of reactants during UOR. 61 Accordingly, the flexible electronic structure of amorphous catalysts contributes to more catalytically active sites for ESM. It is facile to introduce foreign metal/nonmetal elements into amorphous TM-based catalysts. Component regulation usually leads to bond length regulation and electron/charge redistribution, 62 which further optimizes the adsorption/desorption of reaction intermediates and benefits the catalysis process. 63,64 From the crystal models in Figure 2B,C, it can be observed that the atom arrangement of amorphous Fe-doped InPO 4 is disordered compared to the crystal counterpart. Further investigations suggest that the average Fe-O bond length (1.891 Å) in the amorphous Fe-doped InPO 4 is shorter than that of the crystalline one (1.913 Å) ( Figure 2D). A shorter Fe-O bond distance would attribute to more high-valence Fe species and improve the interaction with catalytic intermediates. 44

3.2
Facile structure self-reconstruction TM-based catalysts would undergo in situ structure reconstruction during electrochemical oxidation processes (e.g., OER, UOR, and GOR), resulting in the formation of stable and catalytically active TM (oxy)hydroxide species. 65 Compared with crystalline catalysts, amorphous TM catalysts could easier transform into active TM (oxy)hydroxides and achieve higher catalytic activities. 66,67 For example, Chen  68 Such structure characteristics of CuFeS 2 -BM promote the totally in situ evolution of the original sulfides into metal (oxy)hydroxide phases during OER, whereas electroactive phases/species in the CuFeS 2 -HM catalyst are just confined on the shell/surface ( Figure 3A). Recent studies have taken the advantage of amorphous catalysts' facile structure self-reconstruction feature to design efficient electrocatalysts for ESM. 69,70 In the crystalline NiO/amorphous CrO x heterostructure, the presence of amorphous CrO x can facilitate the selfreconstruction of NiO to the active NiOOH phase and improve adsorption strength toward urea molecules, as evidenced by the in situ electrochemical Raman analysis. 28 The NiO/CrO x catalyst thus shows good UOR performance, and a potential of 1.32 V is acquired to achieve the 10 mA cm −2 with long stability for 5 days. Similarly, a recent report suggests that amorphous Fe-doped NiB (aFe-NiB) needs lower energy (1.40 V vs. reversible hydrogen electrode [RHE]) than the crystalline counterpart (1.45 V vs. RHE) to initiate the structure reconstruction during the UOR process, as illustrated by the in situ Raman spectra ( Figure 3B,C). 66 The aFe-NiB thus exhibits higher UOR performance due to the improved formation of electroactive metal oxyhydroxides.

AMORPHOUS TM ALLOYS
Amorphous TM alloys are highly efficient ESM catalysts due to their high conductivity and the synergistic combination of electroactive metals, as well as rich active sites. 24,71 Co, Ni, and Fe-based amorphous alloys are widely explored ESM electrocatalysts, and current studies mainly focus on nanostructure design and heterostructure construction.

Nanostructure design
Controlling the nanostructure of nanocatalysts is a powerful strategy to upgrade the catalytic performance of amorphous TM alloys by providing a large surface area and promoting the penetration/diffusion of electrolytes. 72,73 Porous materials, 74 nanoarrays, 75 pompoms, 19 films, 76 and aerogels 23 have been investigated for ESM. For instance, Xu et al. used electrodeposition to load amorphous NiFe nanotubes onto the nickel foam (NiFe NTAs-NF) ( Figure 4A). 75 As shown in Figure 4B, the nanotube array nanostructures are presented. The well-defined nanotube arrays offer a large surface area, rich electroactive sites, as well as adequate electrolyte/gas transportation channels during the electrochemical reaction. Together with the amorphous phase of the NiFe alloy, the NiFe NTAs-NF exhibits a high OER activity (η 50 = 216 mV) and good stability for 20 h. Similarly, the amorphous Ni-Co alloy/NF catalyst prepared by a three-step process shows a three-dimensional (3D) hierarchical nanorod array nanostructure ( Figure 4C,D). 77 Compared with the Ni/NF and Co/NF catalysts, the amorphous Ni-Co/NF catalyst shows better performance toward HzOR and attains a higher current density of 1213 mA cm −2 at 0.30 V versus RHE with good stability for 36 h ( Figure 4E,F). Besides nanoarray structure, porous structures also possess a large number of exposed electroactive sites for OER. 78 Typically, the 3D ordered macro-/mesoporous Ni 61 Co 39 alloys with large surface area and distinctive structure show a good performance for OER (η 10 = 241 mV). 74

Heterostructure construction
Besides nanostructure manipulation, constructing heterostructures is another universal strategy to enhance the ESM performance of amorphous TM alloys. Heterostructures can inherit the merits (e.g., high activity, good conductivity, large surface area, and good durability) of different components harmonically and, thereby, contribute to upgraded catalytic performance. Wang et al. developed an eco-friendly exfoliator to assemble amorphous thin layered NiFe alloy nanosheets on two-dimensional (2D) materials (MoS 2 , Ti 3 C 2 MXene, graphene). 24 Taking the NiFe/MoS 2 as an example, the unique 2D/2D configuration largely limits the aggregation and improves the exposure of electroactive sites, and the strong interfacial coupling between MoS 2 and NiFe promotes charge transfer between them. Thus, the NiFe/MoS 2 heterostructure shows a high catalytic activity (η 10 = 260 mV) for OER. Currently, the application of amorphous TM alloys in ESM is still in its infancy, and more studies are encouraged to regulate the physicochemical properties (e.g., porosity, size, conductivity, chemical composition, and shape) of TM alloys to upgrade their catalytic performance.

AMORPHOUS TM (HYDR)OXIDES
TM (hydr)oxides are the most widely explored ESM catalysts owing to their easy preparation, high electrochemical activity, and eco-friendliness. 79

TM oxides
With high redox activity, flexible composition, and strong stability, amorphous TM oxides have aroused great interest in ESM catalysts. 81 Previous achievements suggest two general methods to enhance the catalytic performance of amorphous TM oxides, namely, component regulation, foreign-atom doping, and heterostructure construction.

Component regulation
Compared with single metal oxides, binary/ternary/ quaternary metal oxides with multimetallic active sites and rich redox couples gain growing attention. [82][83][84] A general rule for achieving a high ESM activity is combining more metal species in metal oxides. CoO x is claimed to be more active than FeO x and NiO x , but the presence of a second metal (e.g., Ni, Ce, and Fe) can improve the OER performance of amorphous CoO x significantly. 85 For instance, Pan et al. found that the amorphous Co-Ce oxides (Co 0.9 Ce 0.1 O x ) synthesized by a photochemical metal-organic deposition route show a good OER activity (η 10 = 320 mV). 82 The presence of Ce can significantly improve the catalytic activity of CoO x because of the synergistic effects of suitable metal-OH bond strength, enlarged SSA, and a high Co 3+ proportion. Similar amorphous binary metal oxides such as Fe-Co oxides, 85 Fe-Ni oxides, 86,87 NiWO 4 nanoparticles, 88 and Ni-Co oxides 89 also have been investigated for ESM applications.
To further mediate the multi-electron transfer ESM process, oxides with multiple redox-active metal ion sites are studied. Typically, Fe-Ni-Co oxides that have increased conductivity, populated catalytically active sites, and synergistic action of different elements present as highly efficient catalysts for ESM. 83,90 Aside from the activity, the addition of high-valence state Mo would upgrade Co-Fe oxides' stability. 91 In the amorphous quaternary Ba-Sr-Co-Fe oxide, the presence of abundant coordinately unsaturated active metal sites and the isotropic nature of amorphous oxides contribute to high OER performance. 92

Heteroatom doping
Foreign-atom doping is a universal strategy to adjust the physical/electronic properties of materials. 62 Both cationic and anionic dopings have been performed to optimize the ESM performance of amorphous oxides. Cationic doping generally can improve the charge transfer process, increase ECSA, and optimize the electronic structure of electroactive sites. 93 In an Fe-doped amorphous CoO x , the Fe dopants not only provide active Fe sites to regulate the adsorption/desorption of OER intermediates but also tune neighboring Co sites' electronic structure to increase the OER performance. 26  shows a mostly pH-independent OER activity ( Figure 5C). The pH-dependent OER activity result suggests a lattice oxygen oxidation mechanism (LOM) on the MoCo x O y -100 catalyst. Moreover, tetramethylammonium cation was introduced as a chemical probe to track the existence of superoxo-like (O 2− ) and peroxo-like (O 2 2− ) species that are typical intermediates in LOM-involved OER process. As shown in Figure 5D, the OER performance of MoCo x O y -100 is severely lessened in the presence of tetramethylammonium hydroxide (TMAOH), which reveals the critical contribution of LOM ( Figure 5E).
Anionic doping is also studied to design ESM catalysts, based on amorphous TM oxides. Similar to cationic doping, the addition of anionic dopants can enhance the oxides' activity by regulating the amorphous property and the electronic structure of active sites. In a spinel CuCo 2 O 4 catalyst, the P dopant induces the formation of the amorphous phase that provides rich active sites for OER. 96 The dual anion (N, S)-doping into the amorphous Ni-Co oxide can accelerate the catalytic kinetics by an "ensemble effect," which causes the electron distribution rearrangement. As a result, electrons in active metal centers are migrated into heteroatom, and thus, Co and Ni sites show higher oxidation states for OER. 97

5.1.3
Heterostructure construction Integrating amorphous metal oxides with carbon materials or electroactive materials can enlarge ECSA, populate active sites, and enhance the intrinsic catalytic activity, thereby leading to upgraded catalytic performance. 98 Figure 6A,B). The c-CoNiP x /a-P-MnO y catalyst can achieve the current density of 100 mA cm −2 at 1.35 V versus RHE for UOR, which is 273 mV lower than that for OER ( Figure 6C). In addition, the composite outperforms the single component ( Figure 6D,E), indicating the synergistic effect between the crystalline CoNiP x and amorphous P-MnO y . 102 In the amorphous CrO x confined Ni/NiO nanoparticles catalyst synthesized by a hydrothermal treatment-hydrogenation method ( Figure 6F), the amorphous CrO x can facilitate the in situ generation of active NiOOH phase and the electron transfer process during UOR ( Figure 6G), contributing to high catalytic activities. Additionally, the CrO x shell shows a protective effect and leads to high stability. 103 bimetallic hydroxides via an in situ chemical transformation process. 108 The MOF-derived NiFe-OH-0.75 catalyst with ultrasmall nanoparticles (<10 nm) and abundant catalytically active sites only takes the overpotential of 270 mV to gain 10 mA cm −2 . 2D nanosheets are the most general structure of TM (oxy)hydroxides, which attract great interest. [109][110][111] For example, the porous amorphous nickeliron hydroxide (NiFe(OH) 2 ) nanosheets with multi-metal coordination and highly porous nanostructure exhibit high HzOR activities (j = 10 mA cm −2 @0.06 V vs. RHE) and long-term stability (>22 h). 109 Amorphous nano/microarrays, nanocages, and nanocubes of TM (oxy)hydroxides with unique structures and rich abundance are efficient ESM catalysts. [112][113][114][115] Interestingly, the amorphous Ni-Co hydroxide nanodendrite forests superhydrophilic property, and underwater superaerophobic behavior, which leads to high catalytic performance. 115 This study emphasized that the bubble generation and detachment behavior during the gas-involved reactions would affect the catalytic performance.

Heterostructure construction
Constructing amorphous TM (oxy)hydroxides on the surface of electroactive and highly conductive materials (e.g., TMSs and TMPs) can contribute to high catalytic performance. There are generally two ways to form the amorphous TM (oxy)hydroxides-based heterostructures, presynthesis route, and in situ electrochemical reconstruction. The pre-synthesis route means that amorphous TM (oxy)hydroxides are formed during the catalyst synthesis process via electrodeposition, 116 precipitation, 13,117 chemical deposition, 29 and so on. For instance, Lv et al. prepared crystalline NiFe/amorphous NiFe-(oxy)hydroxide on the NiMo alloy (NM@cNF/aNFO) by electrodeposition, as illustrated in Figure 7A. 118 The abundant nanopores and defects of amorphous NiFe-(oxy)hydroxide, the high  119 Ni-Fe-OH@Ni 3 S 2 /NF, 117 and amorphous NiFe-OH/NiFeP. 116 It is interesting to find that some of the amorphous hydroxide-based composites exhibit high performance toward multiple reactions, which is favorable for system simplification. Li et al. developed a crystalline NiFe phosphide/amorphous P-doped FeOOH hybrid (P-NiFeO x H y ) for UOR, OER, and HER via a corrosion-phosphorization method ( Figure 7B). 120 Compared with the crystalline NiFeO x H y catalyst, P-NiFeO x H y exhibits better performance toward OER and UOR ( Figure 7C) due to its ameliorated adsorption energy of oxygen-containing reactants and superaerophobic and superhydrophilic surface. The P-NiFeO x H y catalyst can be implemented as both anode and cathode for the urea electrolyzer, which can achieve 10 mA cm −2 at a voltage of 1.36 V and could be driven by a solar cell system ( Figure 7D). In alkaline electrolytes, the in situ electrochemical reconstruction mediated TM (oxy) hydroxide formation has been well accepted in the last few years. 121,122 It is suggested that most TM-based catalysts will undergo structure reconstruction and generate (oxy)hydroxides on the catalysts' surface, under alkaline and oxidative conditions. 68,123,124 With this in mind, many studies have focused on the in situ electrochemical construction of high-performance heterostructures for ESM. Yang et al. designed an Ni 0.70 Fe 0.10 V 0.20 S 2 @amorphous NiFe hydroxide catalyst for OER by thermal sulfidation and a subsequent in situ electrochemical treatment ( Figure 7E). 125 The high electronic conductivity of the inner Ni 0.70 Fe 0.10 V 0.20 S 2 , synergistic effects between NiFe hydroxides and Ni 0.70 Fe 0.10 V 0.20 S 2 , and stable electroactive sites on the outer amorphous hydroxide surface together benefit the OER performance (η 10 = 204 mV). In contrast to the pre-synthesis strategy, the inevitable in situ electrochemical conversion process is more convenient and practical for designing TM-based heterostructural catalysts for ESM.

AMORPHOUS TM CHALCOGENIDES
TM chalcogenides (sulfides, selenides, and tellurides) as emerging (pre)catalysts with excellent ESM performance have gained growing scientific attention owing to their high electrical conductivity, good catalytic activity, as well as low cost. 126,127 Typically, amorphous TM chalcogenides with rich defects, dangling bonds, and structural flexibility are highly promising ESM (pre)catalysts. In this section, the recent advances in the development of amorphous TM chalcogenides (sulfides, selenides) for ESM are detailed.

TM sulfides
Amorphous TM sulfides have been extensively used for ESM, especially bimetallic sulfides. 128 Designing amorphous TM sulfides for ESM generally involves the modification of materials' internal and external characteristics, including morphology/nanostructure, electrical conductivity, and electronic structure. 129 To achieve these goals, nanostructure design, heteroatom doping, and heterostructure construction have been extensively studied.

Nanostructure design
To populate the electroactive sites and fasten charge/mass transfer, controlling the nanostructure of amorphous TM sulfides is an efficient method. 2D amorphous TM sulfide nanosheets [130][131][132] and nanoflake arrays 133 are representative catalysts. To avoid the formation of dense structures or films, engineering holes/pores on the sulfide nanosheets can enlarge the surface area of catalysts and promote the release of generated oxygen gas bubbles. 132,134 For instance, the honeycomb-like amorphous Ni-Mo-S membrane composed of abundant intertwined ultrathin nanosheets can realize a current density of 100 mA cm −2 at 1.385 V for UOR. 135 Improving the dispersion of electroactive sites via nanostructure design also contributes to high catalytic performance. Dong et al. prepared highly dispersed amorphous cobalt/nickel sulfides with a nanoflakes array structure by controlling the synthetic conditions. 133 The optimal amorphous Co 4 Ni 1 S nanoflake arrays exhibit good OER and HER activities, and it takes a low cell voltage (1.60 V) to attain 10 mA cm −2 , with high stability (96% activity retention for about 10 h).

Heteroatom doping
Aside from structure regulation, raising the intrinsic activity of amorphous TM sulfides via heteroatom doping is suggested. In 2017, Cai et al. developed oxygen-doped amorphous cobalt sulfide porous nanocubes for OER. 136 The oxygen dopants lead to desirable reaction intermediate (O*) adsorption energy and distorted CoS 4.6 O 0.6 octahedron structure, which synergistically contributes to improved OER activities (η 10 = 290 mV). More recently, ultrathin and amorphous Mo-doped FeS nanosheets with abundant sulfur defects were developed via a hydrothermal route. 137 Benefited from the amorphous ultrathin nanosheets structure and synergistic effects of Mo dopants and S defects, the Mo-doped FeS manifests good OER activities (η 10 = 210 mV, Tafel slope = 50 mV dec −1 ) and stability in the alkaline medium.

Heterostructure construction
Hybridizing amorphous TM sulfides with electroactive nanomaterials (e.g., oxides, metal particles, sulfides, hydroxides, and tellurides) and highly conductive carbon materials can further upgrade the catalytic performance. [138][139][140] Liu et al. designed a bifunctional catalyst composed of Cu cluster-coupled quasi-amorphous on Cu foam (CF) CoS x (Cu@CoS x /CF) via a one-step route ( Figure 8A). 141 With the CF support, the Cu@CoS x /CF is somewhat flexible ( Figure 8B), and the Cu@CoS x /CF manifests a 3D macroporous structure ( Figure 8C). These structural features benefit the exposure of active sites, charge/mass transfer, and mechanical stability. In addition, Cu clusters have multiple effects on the catalytic performance of Cu@CoS x /CF. First, the Cu cluster could improve the electronic transport property of the Cu@CoS x /CF hybrid. Second, the Cu cluster could interact with quasi-amorphous CoS x strongly at the interface, which leads to the generation of rich negative and positive sites and further promotes the interaction of the catalyst with H 2 O. Third, Cu clusters can benefit H 2 O dissociation and facilitate the electrolyzing of water. With these merits, the Cu@CoS x /CF//Cu@CoS x /CF couple can afford 10 mA cm −2 at about 1.50 V in an alkaline electrolyzer. Compositing amorphous TM sulfides with (hydr)oxides/sulfides/tellurides also contributes to high catalytic performance. For example, the NF-supported hierarchical OER catalyst (VO x /Ni 3 S 2 @NF) comprising vanadium oxide nanocrystals (VO x ) embedded in amorphous nickel sulfide (Ni 3 S 2 ) nanosheets can attain 100 mA cm −2 at a low overpotential (358 mV) ( Figure 8D). 142 Apart from the self-supported hierarchical nanostructure and the rational chemical composition, the presence of VO x in the sulfide nanosheets upgrades the catalytic performance by strengthening the Ni-O bonds. Amorphous TM sulfides/chalcogenides heterostructures which gather the advantages of different sulfides and benefit from the synergistic effect among different components also present high ESM activities, such as MoS x @Co 9 S 8 @Ni 3 S 2 /NF, 139 Ni 3 S 2 /MoS x nanosheets/NF, 143 NiTe/amorphous NiS, 34 Mn-Cd-S@amorphous-Ni 3 S 2 , 144 amorphous MoS 2 /CoS/Co 0.85 Se nanotube arrays, 35

TM selenides
Amorphous TM selenides such as Ni x Se y , Fe x Se y , and Co x Se y have attracted great attention as ESM catalysts for their high conductivity, metallic properties, and good electroactivity. 146 35 and amorphous Co-O@Co-Se film 149 are high-performance ESM catalysts. Of note, the amorphous/crystalline heterostructure (Cu-(a-NiSe x /c-NiSe 2 )/TiO 2 ) developed by Park et al. possesses some meaningful characteristics. 37 First, the outer amorphous NiSe x phase can efficiently decrease the size of inner crystalline particles and stabilize them from aggregation, which contributes to a large surface area. Second, the rich defect sites of the amorphous NiSe x provide abundant electroactive centers, and the inner NiSe 2 structure's good electrical conductivity leads to a synergistic effect, thus enhancing the composite's catalytic activity and stability.

TM phosphides and nitrides
TM phosphides and nitrides with metallic properties, high electrical conductivity, and good electrochemical activity are promising ESM catalysts. The design of highperformance amorphous TM phosphides and nitridesbased catalysts generally involves component regulation, nanostructure design, and heterostructure construction.

Component regulation
The introduction of external metals or nonmetals can ameliorate the catalytic performance of single metal phosphides/nitrides. It has been well proved that amorphous NiFeCoP catalysts perform better than their mono/binary metallic counterparts. 150,151 The main reason is that the abundant electroactive sites, disordered atomic arrangements, and synergic effects of different metals contribute to enhanced catalytic activity. A general trend is that more metal species lead to higher ESM catalytic performance, [152][153][154] and the ratio of metals also plays a vital role. 40,155 Aside from the addition of metals, recent studies indicate that anions also can tune the catalytic behavior of amorphous TM phosphides. In a B-involved ternary Co-P catalyst, the Co-P-B could inherit great electron transfer properties from Co-B and Co-P. Hence, the addition of B can regulate the electron transfer process of the ternary catalyst. 156 In addition, the optimal Co-P-B-5 catalyst undergoes obvious structure reconstruction during anodic polarization and facilitates the generation of abundant CoOOH active species on its surface.

Nanostructure design
Catalysts with large particle sizes and dense layer structures are less likely to expose enough electroactive sites for catalytic reactions. In this context, controlling the nanostructure of amorphous TM phosphides/nitrides is in high demand. Currently, amorphous TM phosphides/nitrides with various nanostructures like nanospheres, 157 hierarchical nanosheet arrays, 158 thin films, 159 nanosheets, 160,161 and nanocone arrays 162 have been designed for ESM. An efficient method to construct typical nanostructures is using a template. For example, Huang et al. employed a template-assisted strategy to construct carbon fiber paper-supported amorphous CoMoP x nanosheet arrays (a-CoMoP x /CF) ( Figure 9A). 158 In the synthetic process, Co-MOF and CoMoO 4 act as templates that ease the generation of amorphous CoMoP x by introducing abundant defects and dense voids. Benefiting from the hierarchical nanosheet arrays, amorphous structures, and the synergistic effect of Co/Mo components, the a-CoMoP x /CF shows better activity than the noble metal-based counterparts for full water electrolysis. Employing the a-CoMoP x /CF as the cathode and anode, a current density of 10 mA cm −2 can be attained at an applied cell potential of 1.581 V ( Figure 9B,C), with good stability at 100 mA cm −2 for 100 h ( Figure 9D). Another study used a ZnO template to develop O-incorporated NiMoP on NF (O-NiMoP/NF, Figure 9E), which shows nanotube array architecture with rich pores ( Figure 9F). 163 The O-NiMoP/NF electrode shows good activities toward both UOR and HER, and using O-NiMoP/NF as the bifunctional electrode can realize energy-saving hydrogen production, as schematically in Figure 9G. Compared with the noble metal-based Pt/C/NF||RuO 2 /NF couple, the O-NiMoP/NF electrodes attain a current density of 50 mA cm −2 with much lower energy consumption and a high hydrogen production FE of 96.8% ( Figure 9H,I).

Heterostructure construction
To simultaneously meet the multiple requirements of electrocatalysis, researchers have designed diverse heterostructures based on amorphous TM phosphides/ nitrides. Generally, two types of composites are studied, the amorphous TM phosphides/nitrides-conductive carbon hybrids and amorphous TM phosphides/nitrideselectroactive materials hybrids. Loading amorphous TM phosphides/nitrides on highly conductive carbon materials can decrease particle aggregation and improve electrical conductivity. 41 Zou et al. coupled amorphous CoN x nanoparticles within 3D N-doped graphene aerogel, and the nanohybrid with hierarchical porous structure and rich dual active sites (CoN x and N x C) shows high OER activities (η 10 = 295 mV). 164 In 2020, Li et al. deposited amorphous NiFe phosphides on a 3D N-doped carbon paper (NCP), and the obtained NiFe-P/NCP OER electrode takes a relatively low overpotential of 226 mV to attain 50 mA cm −2 . 165 In these studies, anchoring electroactive species on 3D carbon materials benefits increased ECSA, electron transfer, and mass transportation and, thus, improves the catalytic performance. Interestingly, a recent study suggested that carbon materials can regulate the intrinsic catalytic of amorphous CoP. 166 The halogen (X = F, Cl, and Br)-doped carbon dots (CDs) modified amorphous CoP (X-CDs/CoP) was prepared via a two-step reaction of hydrothermal treatment and phosphorization ( Figure 10A, F-CDs/CoP as an example). Compared with Cl or Br-doped CDs/CoP, the F-doped sample exhibits the best catalytic performance toward OER and HER ( Figure 10B). With good HER and OER performance, the F-CDs/CoP can work as bifunctional catalysts for overall water splitting with high efficiency and good durability over 100 h ( Figure 10C). In-depth density functional theory (DFT) calculations reveal that F-CDs/CoP exhibits lower energy barriers for both HER and OER than those of Cl or Br-doped CDs/CoP ( Figure 10D,E), and the main reason is that the charge transfer between F-CDs and CoP is more prominent compared with the other X-CDs/CoP. Hybridizing amorphous TM phosphides/nitrides with other redox-active materials also gains great interest.

TM borides
Amorphous TM borides are efficient ESM electrocatalysts, which feature facile preparation, high electrical conductiv-ity, and high redox activity. Typically, the reverse electron transfer from B to TMs in amorphous TM borides results in enriched electrons on TMs and thus improves the electrocatalytic process. 72 To engineer high-performance amorphous TM borides for ESM, current efforts focus on nanostructure design, component regulation, heteroatom doping, and heterostructure construction.

Nanostructure design
Amorphous TM borides are generally synthesized by a facile chemical reduction or electroless deposition process, with NaBH 4 or KBH 4 as the reductant and boron source. 171 The violent chemical reaction often leads to severe aggregation of boride particles, which limits the surface area of boride catalysts. To overcome this issue, researchers have developed a series of nanostructures by controlling the reaction conditions (e.g., temperature, time, reactants dosage, solvents, and precursors). For example, Li et al. suggest that CoB x and Ni-FeB nanosheets with a higher surface area show better OER performance than nanoparticle counterparts. 173,174 What is more, the hierarchical structure of mixed metal FeNiCuSnBs guarantees efficient mass/charge transfer, sufficient electrochemical active sites, and rich transportation channels for electrolytes and produced oxygen gas. 175 Recent studies have innovated the design of hierarchically structured amorphous TM borides by using MOF as metal precursors. 176

Heteroatom doping
Heteroatom doping is a universal and favorable method to increase the catalytic activity of amorphous TM borides. 186 In a Ce-doped amorphous NiB catalyst, the synergistic effect of binary Ce and Ni metals contributes to high OER activity. 187 In addition, the self-evolution of the catalyst under the OER condition leads to the formation of CeO 2−x which acts as the electron acceptor and facilitates the generation of crystalline high-valent nickel species, leading to robust electrocatalysis. Besides cationic doping, Chen et al. developed an anion-cation dual doping method to boost OER properties of FeB. The W, P co-doped FeB catalyst was fabricated with a chemical reduction process ( Figure 11A). 42 The W, P co-doped FeB catalyst presents a high OER activity in the alkaline electrolyte (η 10 = 209 mV). It is disclosed that the anion (B and P) etching facilitates the evolution of FeOOH. Further DFT calculations indicate W dopants can boost the intrinsically catalytic activity by improving the conductivity ( Figure 11B) and tuning adsorption/desorption energies of OER intermediates ( Figure 11C). Another study has developed a P-doped NiB for the electrochemical oxidation of HMF. The P dopant leads to relative the displacement of electrons from the surface Ni and increased surface Ni(OH) 2 species, which increases the yield of FDCA from 75.5% of NiB to 90.6% ( Figure 11D). 21

Heterostructure construction
Loading amorphous TM borides on conductive/large area substrates or electroactive materials to construct efficient heterostructures for ESM has been successfully  196 implemented. 188 Substrates like reduced graphene oxide, 189 graphitic carbon nitride, 190,191 black phosphorus sheets, 58,192 and carbon nanofiber [193][194][195] can enhance the catalytic performance of amorphous TM borides. These substrates with large surface areas are deemed to disperse boride particles and thus enlarge the electroactive surface area. In addition, the interaction between borides and substrates can modify borides' electronic structures and upgrade the catalytic property. 193 For instance, the 2D/2D exfoliated black phosphorus (EBP) nanosheet/amorphous CoFeB nanosheets heterostructure (EBP/CoFeB) with wrinkled surface morphology prepared by a chemical reduction method ( Figure 12A,B) manifests high OER activities (η 10 = 227 mV, Figure 12C) with excellent stability in the basic electrolyte. In-depth investigations indicate that the electronic interactions and O affinity difference between CoFeB and EBP can balance the absorption behavior of reaction intermediate for improving the OER and MOR process ( Figure 12D). 58 By tuning the absorption strength of oxygen-containing intermediates, BP-based heterostructures would significantly facilitate the catalytic process. Some studies also have used electroactive materials to support amorphous TM borides, such as Ni hydroxide 196 and pyrite-type boron sulfide. 197 With a chemical surface etching process, the Ni hydroxide phase was formed together with the metallic NiB phase. 196 The obtained F-Ni 2 B/Ni(OH) x catalyst with two phases coexisting on the surface can utilize the abundant adsorption sites in the Ni(OH) x phase as well as the excellent electrical conductivity of the Ni 2 B phase. Thus, the F-Ni 2 B/Ni(OH) x composite outperforms the single Ni 2 B and the noble metal-based RuO 2 catalyst for OER ( Figure 12E,F). In the amorphous FeCoNiB x coated ternary pyrite-type boron sulfide (FeCoNiB x /Ni 0.8 Fe 0.1 Co 0.1 S 2 B x ) catalyst, B atoms reconfigure the electronic structure of metals via downshifting the d-band center and regulating intermediates' binding energies. 197 Moreover, the metallic boride coating enhances the conductivity for OER, and thus, a high activity is attained (η 50 = 392.4 mV).

TM oxometallates
Amorphous TM oxometallates with rich oxygen components such as phosphates, 64 borates, 198 selenites, 199 and silicates 49 are emerging electrocatalysts for ESM. It is suggested that the oxyanions (e.g., PO 4 3− and BO 3 3− ) of TM oxometallates with large negative charges can serve as proton carriers and facilitate the proton transfer process; also, they can stabilize the local pH environment during ESM. [200][201][202] Figure 13A), 48 the sufficient electroactive sites, excellent electrical conductivity, and high structural stability arise from the amorphous-crystalline composite contribute to its good OER performance. Amorphous TM oxometallates/carbon hybrids can be synthesized in a more facile way; as shown in Figure 13B, the Co borate (Co-Bi)/graphene composite was obtained with a chemical reduction process. 214 In the Co-Bi/graphene heterostructure, the graphene supports can not only guide the formation of 2D ultrathin borate nanosheet structure but also expedite electron transfer to promote OER kinetics. Moreover, the synergistic coupled effects between graphene and Co-Bi support charge transport, thereby contributing to superior OER performance in both alkaline and neutral solutions. The structure reconstruction of amorphous TM oxometallates/carbon hybrids has been examined on the amorphous Fe-doped Ni phosphate-carbon nanohybrid (a-NiFePO-C). 215 Compared with its crystalline counterpart, the higher OER activities of a-NiFePO-C are due to its more complete self-reconstruction induced highly electroactive Fe(III)-NiOOH-carbon structures ( Figure 13C), and low charge-transfer resistance.

CONCLUSIONS AND PERSPECTIVES
Designing efficient amorphous low-cost ESM catalysts is an urgent task in advancing hydrogen energy. This review provides a critical analysis of recent achievements in the design of amorphous TM-based catalysts for ESM. The main features of amorphous catalysts include large SSA, flexible electronic structure, and deep structure reconstruction. Powerful catalyst design strategies such as nanostructure design, component regulation, heteroatom doping, and heterostructure construction have been successfully implemented to design efficient TM alloys, (hydr)oxides, chalcogenides, phosphides, nitrides, borides, and oxometallates for ESM (Scheme 2). These strategies can efficiently improve the catalytic performance of amor-phous materials by providing more electroactive sites, ensuring fast mass/charge transfer, enlarging ECSA, regulating catalysts' electronic properties, and improving the structure stability. Despite current achievements, some critical issues deserve further study: 1. Developing new techniques to realize the large-scale production of low-cost amorphous electrocatalysts is highly urgent. Reducing amorphous TM-based electrocatalysts' fabrication costs is an urgent mission, which would largely improve their cost-effectiveness. Given the environmental impact of catalyst preparation, it is favorable to employ methods with low energy consumption and/or carbon emissions such as chemical precipitation, electrodeposition, flash Joule heating, ball milling, and plasma synthesis. In addition, it is necessary to design protocols for large-scale catalyst preparation to meet the demand of industrial applications. 2. Innovating novel catalyst design strategies to enhance catalytic performance is necessary. Compared with one S C H E M E 2 A summary of key strategies for amorphous electrocatalyst design.
individual strategy (e.g., doping, nanostructure control, heterostructure construction, or component optimization), the co-implementation of multiple strategies would combine the advantages of individuals and lead to optimal catalysts. Considering the rapid development of single-atom catalysts, combining single/dualatom catalysts with amorphous TM nanomaterials might lead to highly efficient catalysts for ESM. Moreover, by combining the catalytic properties of different metals, the development of amorphous high entropy catalysts would generate novel catalyst systems. 3. It is suggested to utilize different advanced techniques (e.g., X-ray absorption spectroscopy, high-resolution transmission electron microscopy, and DFT calculations) to figure out the fine structure of amorphous catalysts, which could assist the interpretation of the structure-performance relationship and guide future catalyst design. Of note, it is still challenging to determine the exact structure of amorphous catalysts due to the feature of long-range disorder, which significantly hinders the computations of amorphous catalysts. In this context, it is necessary to figure out the structure of amorphous materials with advanced X-ray techniques (e.g., X-ray absorption spectroscopy), atomic electron tomography, fluctuation electron microscopy, and so on before performing DFT calculations and molecular dynamics simulations. Another issue is that most TM-based catalysts undergo in situ structure selfreconstruction during electrooxidation processes, and thus, it is important to track catalysts' structure selftransformation and understand the dynamic evolution of electroactive sites. 4. Implementing amorphous catalysts in industrial applications is necessary for their further applications. Currently, most electrochemical tests are performed on a laboratory scale, which is far from industrial conditions. Accordingly, measuring catalysts' activities at high current densities and long testing periods would help to take catalysts' application one step further. This requires the development of high-performance catalysts and large-scale electrochemical reactors. Aside from water electrolysis that only produces highpurity gaseous products, the electrolysis of alcohols and complex biomass will generate value-added liquid products that require separation and purification facilities. 5. Aside from electrochemical oxidation of small molecules to produce value-added chemicals/fuels, it is significant to apply amorphous catalysts in other fields to address pressing environmental and energy challenges, such as inorganic/organic pollutant oxidation/reduction, oxygen reduction, carbon dioxide/carbon monoxide reduction, nitrogen/nitrate reduction, hydrogen oxidation, and biowaste conversion. With high redox properties and low cost, amorphous TM-based catalysts would help to improve the catalytic reaction efficiency and save energy input for different applications.

C O N F L I C T O F I N T E R E S T S TAT E M E N T
The authors declare no conflicts of interest.