Amorphous nanomaterials for energy catalysis: a review

Reducing agent regulation method

The reducing agent regulation method represents a core strategy for amorphous synthesis based on wet-chemical systems. Its underlying formation mechanism involves selecting an appropriate reducing agent to precisely control the reduction rate of precursor ions. The kinetic imbalance between the rate of atom generation and the combined rates of atomic diffusion and ordered arrangement, suppressed long-range crystalline periodicity while promoting the assembly of atoms into a short-range ordered, long-range disordered configurations, ultimately yielding an amorphous structure. This approach does not require extreme reaction conditions and enables precise control over the phase, morphology, and size of amorphous materials through regulation of the type and amount of reducing agent, as well as the reaction environment. It is a commonly employed synthesis route for metal-based and metal oxide-based amorphous catalytic materials^[59,60]. Duan et al.^[43] achieved the facile synthesis of both amorphous and crystalline copper nanoparticles in an aqueous solution system at room temperature using copper chloride as the copper source [Figure 2A]. Using deionized water as a green liquid-phase medium without the addition of surfactants, templating agents, or dispersants, and employing tannic acid as a mild reducing agent to create a gentle reduction environment, Cu²⁺ was slowly reduced to Cu⁰ atoms. Under these conditions, the reduction rate was significantly lower than the atomic diffusion rate, preventing copper atoms from spontaneously arranging in an ordered manner to form a crystalline lattice under thermodynamic driving forces. Instead, they nucleated and grew in a short-range ordered, long-range disordered fashion, ultimately forming amorphous copper nanoparticles with an average particle size of 3.33 nm with excellent dispersion [Figure 2B]. Conversely, when the reducing agent was replaced by sodium borohydride, its strong reducing properties led to the rapid reduction of Cu²⁺, generating a large number of Cu⁰ atoms instantaneously^[61]. These atoms quickly aggregated and spontaneously crystallized under thermodynamic stabilization tendencies, forming well-crystallized crystalline copper nanoparticles. This entire process was conducted at room temperature and atmospheric pressure, requiring no specialized equipment or harsh reaction conditions. The as-prepared amorphous copper nanoparticles exhibited a large electrochemical active surface area and enhanced CO₂ adsorption capacity. When employed as electrocatalysts for CO₂ reduction, the amorphous copper nanoparticles demonstrated significantly superior electrocatalytic performance compared to their crystalline counterparts, achieving Faradaic efficiencies of 37% for formate and 22% for ethanol at -1.4 V, a maximum total Faradaic efficiency of 59% for liquid products, along with good catalytic stability over 12 h. This synthetic strategy is also applicable for preparing amorphous metal oxide catalytic materials. In the synthesis of boron-doped amorphous CeO₂/GO composite materials^[46], the researchers utilized sodium borohydride as both a boron source, reducing agent, and precipitant. Through a wet-chemical method^[11], boron was uniformly doped into the CeO₂ lattice. The introduction of boron atoms formed planar triangular BO₃³⁺ structural units, disrupting the intrinsic Fm-3m high-symmetry crystal structure of CeO₂, inducing symmetry breaking and amorphization [Figure 2C]. Simultaneously, the additional electrons introduced by the boron atoms effectively promoted the reduction of Ce⁴⁺ to Ce³⁺, ultimately enabling the controllable preparation of an amorphous CeO₂/GO composite material with a high Ce³⁺ content (85.7%) [Figure 2D]. To elucidate the influence of the amorphous structure on material properties, the study also synthesized a purely crystalline CeO₂/GO control sample using a glycine-sodium hydroxide buffer solution instead of NaBH₄ [Figure 2E], directly confirming the crucial role of the pure amorphous structure in enhancing CO₂ photoreduction performance. The amorphous CeO₂/GO composite exhibited excellent CO production performance in dilute CO₂ feed streams through photoreduction, fully demonstrating the efficacy of the reduction regulation strategy in constructing amorphous metal oxide phases and investigating structure-property relationships.

Figure 2. (A) Schematic representation of a-Cu and c-Cu formation; (B) TEM image of a-Cu. Figure 2A and B is reprinted with permission from^[43]. Copyright 2018, John Wiley and Sons; (C) Structural evolution from pristine CeO₂ to B-doped CeO₂ (B-7.69%) upon boron incorporation; (D) TEM image of the B-doped CeO₂/GO nanosheet, with the insets displaying the selected area electron diffraction (SAED) pattern and high-resolution transmission electron microscopy (HRTEM) image; (E) Schematic of the synthesis procedure for B-doped CeO₂/GO nanosheet. Figure 2C-E is reprinted with permission from^[46]. Copyright 2025, John Wiley and Sons. PVP: Polyvinyl pyrrolidone; TEM: transmission electron microscopy; GO: graphene oxide.

Heat treatment

The low-temperature annealing induction method is currently the most widely employed post-treatment strategy for crystal phase regulation in amorphous inorganic catalytic materials. As an important branch of solid-state amorphization techniques, it is often combined with classical precursor preparation routes such as liquid-phase impregnation, sol-gel processing, hydrothermal synthesis, and chemical precipitation. Its core design principal leverages annealing temperature as the key regulation parameter, strictly controlling the thermal treatment temperature below the critical crystallization temperature of the target material. By leveraging the kinetic inhibition effects of low-temperature environments on atomic migration, rearrangement, and crystalline phase evolution, this method suppresses crystal nucleation and grain growth, stabilizing the system in a short-range ordered, long-range disordered amorphous state. Simultaneously, it achieves uniform dispersion and surface anchoring of active components, establishing itself as a versatile and reliable technical pathway for efficiently constructing amorphous-based composite catalytic materials. For the vast majority of inorganic catalytic materials, transitioning from an amorphous precursor to a crystalline phase requires overcoming a specific energy barrier. Only upon reaching the critical crystallization temperature do atoms acquire sufficient thermal kinetic energy to undergo long-range diffusion and ordered lattice assembly, thereby enabling nucleation and crystal growth. Under low-temperature annealing conditions, the system facilitates precursor compositional transformations - such as the conversion of hydroxides, chlorides, or nitrates into catalytically oxides or sulfides - while remaining energetically insufficient to induce long-range ordered atomic arrangements, thus stabilizing the material in an amorphous metastable state^[62]. Conversely, when the temperature exceeds the critical crystallization temperature, atomic thermal motion intensifies significantly, leading to directed arrangement, and the material gradually crystallizes, forming a complete lattice structure^[63,64]. Based precisely on this temperature-dependent phase transformation behavior, the low-temperature annealing induction method enables controllable switching between amorphous and crystalline phases using the same chemical composition and precursor system, providing an ideal research paradigm for establishing structure-property relationships in catalytic materials. For example, Gao et al.^[44] used hydrothermally synthesized flower-like anatase TiO₂ nanosheets as a support. After loading a RuCl₃·xH₂O precursor via liquid-phase impregnation, they subjected the material to low-temperature annealing at 100 °C. Utilizing the kinetic inhibition of RuO₂ crystallization at low temperatures, they obtained an a-RuO₂/TiO₂ composite catalyst with uniformly loaded amorphous RuO₂. By increasing the annealing temperature to 600 °C, they obtained a crystalline c-RuO₂/TiO₂ control sample using the same precursor and support system, achieving precise and controllable regulation of the RuO₂ crystal phase through different annealing temperatures [Figure 3A]. Characterization techniques such as electron diffraction [Figures 3B and C] clearly revealed the absence of RuO₂-related diffraction rings or spots in a-RuO₂/TiO₂, confirming the amorphous nature of RuO₂. In contrast, distinct RuO₂ characteristic diffraction rings or spots were observed in c-RuO₂/TiO₂, confirming its crystalline structure. Electrochemical tests demonstrated that the a-RuO₂/TiO₂ prepared by low-temperature annealing exhibited superior cathode catalytic performance in lithium-oxygen batteries, achieving a discharge specific capacity of 2.3 mAh cm^-2 at a current density of 0.04 mA and maintaining stable cycling for over 160 cycles, far outperforming c-RuO₂/TiO₂ [Figure 3D]. This performance advantage originates from the unique structural characteristics engineered by low-temperature annealing: suppressing of crystallization concurrent with retention of a high specific surface area (131 m² g^-1), a high density of surface defects and abundant surface functional groups, providing ample active sites for oxygen reduction and evolution reactions. Furthermore, the amorphous catalyst induced the formation of 100-300 nm amorphous Li_2-xO₂ discharge products [Figures 3E and F], which could be completely decomposed during the charging process, significantly enhancing the reversibility of the oxygen electrode reaction and charge transfer kinetics. Concurrently, Shi et al.^[45] applied the low-temperature annealing induction method to prepare ultrathin amorphous CoO nanosheets (a-CoO) for photocatalytic CO₂ reduction. Using Co(acac)₂ and NaNO₃ as precursors, they directly synthesized amorphous CoO nanosheets via low-temperature annealing at 230 °C under an argon atmosphere. Upon further increasing the annealing temperature to 400 °C under identical atmosphere conditions, a crystalline CoO (c-CoO) reference sample was obtained [Figure 3G]. Photocatalytic CO₂ reduction performance tests revealed that the a-CoO prepared by low-temperature annealing exhibited significantly higher catalytic activity than the crystalline sample and maintained essentially unchanged catalytic performance over five cycles. The origin of this performance advantage is attributed to the amorphous structure formed by low-temperature crystallization inhibition, which induced a transition in the local coordination environment of Co atoms from quasi-octahedral to quasi-tetrahedral. This regulation altered the energy level distribution of Co 3d orbitals and increased the number of unpaired electrons in the frontier d orbitals, significantly enhancing d-p orbital hybridization between Co 3d and CO₂ 2p orbitals, thereby lowering the CO₂ activation energy barrier [Figure 3H]. In contrast, the high-temperature crystallized cubic CoO exhibits significantly weakened adsorption and activation capabilities toward CO₂, resulting in a substantial decline in photocatalytic performance. Characterized by its distinct temperature dependence, method compatibility, structural retention, and comparative structure-property analysis capabilities, the low-temperature annealing induction method transcends the limitations of traditional amorphous preparation methods that rely on extreme conditions. It simplifies crystal phase regulation from complex synthetic design to a quantifiable and reproducible temperature control, establishing itself as a representative and widely adopted standard synthesis strategies in the field of amorphous catalytic materials.

Figure 3. (A) Schematic depiction of the synthesis procedure for a-RuO₂/TiO₂ and c-RuO₂/TiO₂ catalysts. Electron diffraction pattern of (B) a-RuO₂/TiO₂ and (C) c-RuO₂/TiO₂; (D) Terminal voltage profiles of a-RuO₂/TiO₂ and c-RuO₂/TiO₂ electrodes at 0.04mA. Li 1s XPS spectra of (E) a-RuO₂/TiO₂ and (f) c-RuO₂/TiO₂ electrode in the fully discharged state. Figure 3A-F is reprinted with permission from^[44]. Copyright 2023, American Chemical Society; (G) Synthetic schematic of a-CoO and c-CoO; (H) Free energy diagrams and corresponding configurations for CO₂ photoreduction processes (color code: Co blue, O red, H white, C brown). Figure 3G and H is reprinted with permission from^[45]. Copyright 2025, John Wiley and Sons. XPS: X-ray photoelectron spectroscopy.

Despite the remarkable intrinsic activity offered by pure amorphous phases, their practical application is often hindered by insufficient electronic conductivity and structural instability under harsh operating conditions. To address these limitations, researchers have advanced beyond single-phase architectures to develop amorphous/crystalline heterojunction catalysts, specifically engineered to synergistically integrate the complementary functional advantages of both structural motifs.

Top-down approaches

The fundamental principle of top-down strategies lies in utilizing external stimuli to disrupt the single-phase equilibrium of a precursor, inducing a phase transformation in localized regions while preserving portions of the original phase, thereby forming an amorphous/crystalline heterojunction. Based on the type of stimulus, these strategies can be further subdivided into chemical reduction-induced and physical annealing-induced processing routes. Chemical reduction-induced strategies have garnered widespread attention due to their ability to precisely control the degree of phase transformation and interfacial characteristics under mild conditions. The core mechanism of this approach involves selective reactions between reducing agents and precursors, which disrupt the long-range ordered structure in specific regions or induce crystallization of particular components, while simultaneously introducing defects (such as oxygen vacancies) and reconstructing electronic structures, ultimately forming heterointerfaces that combine the advantages of both phases. The success of this method hinges on precise control of reducing agent dosage, reaction temperature, and duration. Only an appropriate degree of reduction yields heterointerfaces with both high interfacial density and structural stability.

Wang et al.^[49] used NiCo-metal–organic framework (NiCo-MOF) grown in situ on carbon cloth as a precursor, which was calcined in air to obtain crystalline NiCo₂O₄ nanosheet arrays. Subsequently, the crystalline NiCo₂O₄ was reduced using a NaBH₄ solution at room temperature, ultimately yielding an amorphous/crystalline heterojunction NiCo₂O₄-B-CC electrode supported on carbon cloth [Figure 4A]. The strongly reducing H^- ions provided by NaBH₄ acted as oxygen scavengers, not only reducing the crystallinity of NiCo₂O₄ and inducing a partial transformation to an amorphous phase, but also successfully introducing a large number of oxygen vacancies, forming a well-defined amorphous/crystalline heterointerface [Figure 4B]. This gradual amorphization from the exterior inward preserved the high conductivity of the crystalline core while significantly increasing the density of active sites for both the oxygen evolution reaction (OER) and hydrogen evolution reaction (HER) through the oxygen vacancy defects in the amorphous shell. Consequently, NiCo₂O₄-B-CC exhibited excellent performance in electrocatalytic water splitting. Density functional theory (DFT) calculations revealed that the incorporation of oxygen vacancies induces a pronounced rearrangement of the local coordination environment surrounding the catalytically active centers. Such structural modulation significantly attenuates the free energy barriers corresponding to the potential-determining steps for both the OER and HER processes. In contrast, Fan et al.^[75] developed another reduction-induced pathway. They first synthesized amorphous Ni_0.5Fe_0.5Mo_1.5O_x ternary oxide nanosheets via a solvothermal method. Subsequently, they added varying amounts of borane tert-butylamine complex in a triethylene glycol system at 250 °C for reduction treatment [Figure 4C]. With an optimal amount of reducing agent, partial Ni and Fe ions were reduced and precipitated in situ, forming crystalline Ni and FeNi₃ nanoparticles with a mean dimension of 3.9 nm [Figure 4D-F], uniformly anchored onto the residual amorphous oxide matrix. This inward precipitation mechanism resulted in the formation of numerous intimately contacted heterointerfaces between the crystalline metal nanodomains and the amorphous oxide. X-ray photoelectron spectroscopy (XPS) revealed a clear electronic coupling effect. The resulting material exhibited a low OER overpotential of 278 mV and operated via a lattice oxygen-mediated reaction mechanism. This study systematically demonstrates that the degree of crystalline phase precipitation and interface structure can be precisely regulated through controlled modulation of the reducing agent dosage. Insufficient reducing agent promotes selective Ni precipitation only, whereas excess dosage resulted in extensive alloying or even structural collapse, highlighting the critical importance of moderate chemical reduction in constructing high-performance heterointerfaces.

Figure 4. (A) Schematic illustration of the synthesis process of the NiCo₂O₄-B-CC; (B) HRTEM images of NiCo₂O₄-CC. Figure 4A and B is reprinted with permission from^[49]. Copyright 2024, John Wiley and Sons; (C) Schematic illustration showing the synthesis of typical Ni-FeNi₃/Ni_0.5-bFe_0.5-yMo_1.5O_x nanohybrids; (D) High-magnification TEM image; (E and F) HRTEM images of Ni-FeNi₃/Ni_0.5-bFe_0.5-yMo_1.5O_x. Figure 4C-F is reprinted with permission from^[75]. Copyright 2023, John Wiley and Sons. BTBC: Borane-tert-butylamine complex; CC: carbon cloth; HRTEM: high-resolution transmission electron microscopy.

The annealing-induced strategy constitutes a core implementation approach for top-down in situ partial crystallization. It is grounded in the use of an amorphous material as the single precursor. Through precisely controlled thermal energy input, this strategy disrupts the thermodynamic equilibrium associated with the disordered atomic arrangement, thereby exploiting differences in crystallization energy barriers among different structural regions or chemical components. As a result, selective crystallization is achieved locally, while retaining a portion of the original amorphous phase. This enables the in-situ construction of structurally dense amorphous/crystalline heterojunction with significant synergistic effects. Compared to other top-down strategies, this method does not require the introduction of additional chemical reagents; phase transformation control is achieved solely through thermodynamic regulation, simultaneously optimizing interfacial bonding strength, defect density, and electronic structure within the heterojunction. Key process control parameters include annealing temperature, holding time, heating rate, and atmospheric environment, with annealing temperature serves as the predominant factor determining the degree of phase transformation^[76]. However, excessively high temperatures can result in complete crystallization of the amorphous phase or collapse of the heterojunction. Therefore, precise temperature control is essential for optimizing the ratio and synergistic coexistence of amorphous and crystalline phases. For instance, Zhao et al.^[77] first synthesized amorphous RuSe₂ nanoparticles with a porous structure using F127 as a soft template via a simple hydrothermal method, followed by annealing treatments at different temperatures under an argon atmosphere [Figure 5A]. Structural characterization revealed that the as-prepared pristine sample was completely amorphous. After annealing at 300 °C, the material remained predominantly amorphous. However, upon increasing the annealing temperature to 400 °C, distinct RuSe₂ lattice fringes began to appear, alternating with amorphous regions, forming abundant amorphous/crystalline interfaces [Figure 5B]. Further increasing the temperature above 500 °C led to complete crystallization and the disappearance of amorphous regions. Electrochemical tests showed that RuSe₂ annealed at 400 °C (RuSe₂-400), featuring coexisting amorphous and crystalline phases, required an overpotential of only 27 mV to achieve a current density of 10 mA cm^-2 in the alkaline HER, significantly outperforming the fully amorphous or fully crystalline counterparts. DFT calculations revealed that the amorphous structure facilitates the adsorption and activation of water molecules (adsorption energy -2.84 eV), while the crystalline structure optimizes the adsorption free energy of hydrogen intermediates (-0.19 eV). The synergistic effect of these two phases makes the amorphous/crystalline interface an ideal active center for the HER. Similarly, Zhang et al.^[78] adopted a facile Pechini method combined with gradient annealing in an air atmosphere to precisely regulate the crystallization degree by tuning the annealing temperature. An optimal amorphous/crystalline heterojunction was constructed at 300-350 °C, while complete crystallization occurred when the temperature exceeded 400 °C [Figure 5C]. The MnRuO_x-300 sample obtained by heat treatment at 300 °C presents a solid-solution structure with RuO₂ microcrystals embedded in an amorphous MnRuO_x matrix [Figure 5D]. The amorphous/crystalline heterointerface can induce abundant oxygen vacancies [Figure 5E], enlarge the electrochemically active surface area, and accelerate the interfacial charge transfer kinetics. DFT calculations further verify that the amorphous/crystalline heterojunction can effectively regulate the d-band center, reduce the energy barrier of the OER [Figure 5F], and strengthen the interfacial charge transfer behavior [Figure 5G]. Benefiting from the interfacial synergistic effect between RuO₂ microcrystals and amorphous MnRuO_x, the catalyst achieves simultaneous enhancement of acidic OER catalytic activity and structural stability. It delivers an extremely low overpotential and Tafel slope, and exhibits outstanding long-term operational stability under high current densities in both acidic electrolytes and proton exchange membrane water electrolyzer.

Figure 5. (A) Schematic diagram of the synthetic procedure for amorphous porous RuSe₂; (B) HRTEM image of RuSe₂-400 ℃. The inset (B) shows its corresponding FFT pattern. Figure 5A and B is reprinted with permission from^[77]. Copyright 2024, Elsevier; (C) The schematic illustration of MnRuO_x catalyst preparation; (D) HAADF-STEM image of MnRuO_x-300; (E) The O 1s XPS spectra of MnRuO_x-300 and MnRuO_x-500 catalysts; (F) The OER energy profile at an applied potential of 1.23 V. (G) Charge density difference of C/C-MnRuO_x and C/A-MnRuO_x. Figure 5C-G is reprinted with permission from^[78]. Copyright 2024, John Wiley and Sons. HRTEM: High-resolution transmission electron microscopy; XPS: X-ray photoelectron spectroscopy; OER: oxygen evolution reaction; HAADF-STEM: high-angle annular dark-field scanning transmission electron microscopy.

Bottom-up approaches

Within the synthetic methodology for amorphous/crystalline heterojunction, the bottom-up construction approach constitutes a fundamental strategy. In contrast to top-down partial crystallization, the construction approach follows the logic of “pre-fabricating the substrate, then loading the secondary phase”. This approach entails the prior synthesis of a substrate material with a well-defined crystalline or amorphous structure, followed by the controlled introduction of the complementary phase onto its surface through epitaxial growth, solvothermal deposition, or electrochemical assembly, thereby forming an amorphous/crystalline heterojunction at the interface. The core merit of this strategy stems from the programmability of interface design. By independently controlling the structure, composition, and morphology of both the substrate and the loaded phase, it enables precise tailoring of interface density, interface energy level alignment, and electronic interactions. Based on the combination mode of the substrate and the loaded phase, the stepwise construction method can be further subdivided into several models, including loading an amorphous phase onto a crystalline substrate, loading a crystalline phase onto an amorphous substrate, and polymorphic epitaxy.

As a representative example of loading an amorphous phase onto a crystalline substrate. Li et al.^[47] first grew Ni(OH)₂ nanosheet arrays on nickel foam via a hydrothermal method, which were subsequently converted to crystalline NiO nanosheets by annealing at 300 °C under argon. This crystalline substrate was then immersed in a RuCl₃ solution for room-temperature etching, allowing Ru³⁺ ions to adsorb onto the NiO surface via electrostatic interactions and partially substitute Ni sites. Finally, a secondary annealing step at 300 °C under argon thermally converted the adsorbed Ru species into amorphous RuO₂, thereby establishing a core-shell heterojunction comprising an amorphous RuO₂ shell encapsulating a crystalline NiO core (a-RuO₂/NiO) [Figure 6A]. High-resolution transmission electron microscopy (HRTEM) clearly showed well-defined NiO lattice fringes uniformly coated by an amorphous RuO₂ layer, with a tightly bonded interface and no apparent lattice mismatch. Electron paramagnetic resonance (EPR) and XPS revealed that the introduction of amorphous RuO₂ significantly increased the oxygen vacancy concentration, thereby enhancing the spin-polarized electron density of the material. This catalyst required an overpotential of only 25 mV to achieve a current density of 10 mA cm^-2 in the alkaline HER and exhibited excellent bifunctional catalytic performance for both the HER and the urea oxidation reaction (UOR). DFT calculations revealed that the electron redistribution at the amorphous/crystalline interface optimized the d-band center position, significantly reducing the adsorption energies of H₂O and CO intermediates, thereby accelerating the reaction kinetics. In contrast to the above model, Qian et al.^[79] used nickel-iron bimetallic foam (NIF) as a substrate. First, amorphous Ni_xFe_yS microspheres assembled from nanosheets were grown on the NIF surface via a two-step solvothermal method. Subsequently, a second solvothermal treatment in a solution containing Co²⁺ and phosphotungstic acid led to the uniform growth of crystalline CoWO₄ nanoparticles on the surface of the amorphous Ni_xFe_yS microspheres, forming an amorphous/crystalline heterojunction microsphere [Figure 6B]. X-ray diffraction (XRD) and selected area electron diffraction (SAED) confirmed the good crystallinity of CoWO₄, while Ni_xFe_yS remained amorphous, with a clear interface between them. XPS showed shifts in the Co 2p and Fe 2p peak positions after the formation of the heterointerface, indicating electronic interactions at the interface. The amorphous Ni_xFe_yS shell effectively protected the internal sulfur components from excessive dissolution, while the crystalline CoWO₄ partially transformed into highly active cobalt oxyhydroxides during the reaction, synergistically maintaining catalytic activity and structural stability. This catalyst required an overpotential of only 322.8 mV to drive a high current density of 500 mA cm^-2 for the OER and 306.5 mV for the HER, demonstrating excellent bifunctional water-splitting performance. Notably, the electrode operated stably for over 48 hours under industrial conditions using 30 wt% KOH electrolyte. The stepwise construction method can also achieve polymorphic epitaxial growth of amorphous and crystalline phases. Li et al.^[48] used nickel foam as a substrate to first synthesize Pt-doped amorphous (Co, Ni)₁₁(HPO₃)₈(OH)₆ nanowires (Pt-a-CoNiHPi) via a two-step solvothermal method. Subsequently, during a second solvothermal process, the remaining reactive sites on the amorphous surface induced the epitaxial growth of crystalline (Co, Ni)₁₁(HPO₃)₈(OH)₆ nanocones (Pt-c-CoNiHPi), ultimately forming a polymorphic Heterojunction with amorphous nanowires encapsulating crystalline nanocones [Figure 6C]. Inductively coupled plasma optical emission spectrometry (ICP-OES) determined a Pt doping content of only 0.21 wt% (3.31 μg cm^-2). DFT calculations showed that the substitution energy for Pt in the amorphous phase (0.558 eV) was significantly lower than that in the crystalline phase (0.924 eV), confirming that the flexible nature of the amorphous structure is more favorable for noble metal doping. After the formation of the heterogeneous interface, electrons transfer from Ni/Co to Pt, rendering Pt in a partial zero-valent state and increasing the binding energies of Ni and Co. This catalyst exhibited an exceptionally low overpotential of 19 mV for the alkaline HER at 10 mA cm^-2 and an outstanding mass activity of 39.1 mA μg^-1, significantly surpassing that of commercial Pt/C. Theoretical calculations further revealed that the electron redistribution at the amorphous/crystalline interface optimized the d-band center position of Pt, bringing the hydrogen adsorption free energy close to zero (0.12 eV), while simultaneously significantly reducing the adsorption energy of water molecules at Ni sites. These synergistic effects collectively accelerated the kinetics of the Volmer-Tafel reaction pathway.

Figure 6. (A) Schematic illustration of the synthesis of a-RuO₂/NiO nanosheets supported on nickel foam. Reprinted with permission from^[47]. Copyright 2023, American Chemical Society; (B) Schematic diagram of the formation process of CoWO₄/Ni_xFe_yS/NIF via two-step solvothermal method. Reprinted with permission from^[79]. Copyright 2024, Elsevier; (C) Schematic diagram of synthesis and structure of Pt-a/c-NiCoHPi. Reprinted with permission from^[48]. Copyright 2023, John Wiley and Sons. NIF: Nickel-iron bimetallic foam.

Supersaturated co-precipitation method

The supersaturated co-precipitation method combines the synthesis of the amorphous support with the loading of single atoms into a single step. Unlike traditional multi-step syntheses, this method utilizes fast reaction kinetics to disrupt the thermodynamic equilibrium state of atomic ordering, allowing guest metal atoms to be directly encapsulated into the framework of the amorphous matrix at the moment of its formation. This not only simplifies the synthesis procedure but, more importantly, achieves uniform doping with high single-atom loadings, overcoming the limitation of crystalline supports regarding guest metal atom loading. Li et al.^[53] successfully synthesized a Ru single-atom anchored amorphous NiMoO_x catalyst using a one-step supersaturated co-precipitation method. They rapidly mixed high-concentration precursor solutions of Ni²⁺, MoO₄^2-, and Ru³⁺, inducing instantaneous co-precipitation with the assistance of ultrasonication. Due to the extremely fast precipitation rate, atoms did not have time to undergo long-range ordering, directly forming a structure with Ru atoms uniformly dispersed within the amorphous NiMoO_x network [Figure 7A]. During this process, Ru atoms were directly “frozen” into the amorphous framework, achieving uniform doping at the atomic scale [Figure 7B]. Structural characterization revealed that the M-O bond lengths in amorphous NiMoO_x were longer than those in crystalline NiMoO₄, indicating a looser amorphous structure conducive to exposing more active sites. Meanwhile, X-ray absorption spectroscopy confirmed electronic interactions between Ru and Ni, and this optimized electronic structure endowed the catalyst with excellent activity and stability in the OER.

Figure 7. (A) Schematic view for preparing amorphous catalysts by one-step co-precipitation strategy; (B) AC-HAADF-STEM image of the a-RNMO sample. Figure 7A and B is Reprinted with permission from^[53]. Copyright 2025, Springer Nature; (C) Schematic depiction of the fabrication of AeZrO₂ and the subsequent loading of Pt single atoms. Reprinted with permission from^[86]. Copyright 2022, Elsevier; (D) Schematic diagram of the synthetic procedure for Pt/TiB_xO_y. Reprinted with permission from^[51]. Copyright 2022, American Chemical Society. PVP: Polyvinyl pyrrolidone; AC-HAADF-STEM: aberration-corrected high-angle annular dark-field scanning transmission electron microscopy; a-RNMO: Ru single atoms onto amorphous NiMoO.

Thermal-driven method

The thermal-driven method represents another important synthesis strategy. Its core principle involves leveraging thermal annealing to either activate pre-existing intrinsic defects within the amorphous support or create new defects and coordination sites in situ during the amorphization process, thereby providing the driving force and stable environment for single-atom anchoring. Dong et al.^[86] successfully anchored Pt single atoms onto amorphous ZrO₂ nanowires via an impregnation-reduction method. They first synthesized amorphous ZrO₂ nanowires rich in oxygen vacancies using a wet-chemical method. Subsequently, the support was immersed in an H₂PtCl₆ solution, allowing Pt ions to adsorb near the oxygen vacancies. Finally, thermal reduction was performed at 200 °C under an H₂/Ar atmosphere [Figure 7C]. During this process, thermal annealing not only reduced Pt ions to the single-atom state but, more importantly, activated the oxygen vacancy defects on the surface of the amorphous ZrO₂, enabling them to act as “chemical traps” to form stable Pt-O-Zr bridging bonds with Pt atoms. X-ray absorption fine structure (XAFS) spectroscopy confirmed that Pt atoms were dispersed as single atoms and formed Pt-O coordination with the support. This strategy fully leveraged the intrinsic defects inherent to amorphous materials, activating them as anchoring sites for single atoms through mild heat treatment. In contrast to the above strategy, Cheng et al.^[51] developed a more proactive thermal-driven strategy involving high-temperature reduction treatment to introduce heteroatom defects in situ concurrently with the amorphization of the support, thereby creating a novel coordination environment for single atoms. Using crystalline Pt/TiO₂ as a precursor, they subjected it to reduction treatment with NaBH₄ at high temperature [Figure 7D]. During this process, B atoms generated from the decomposition of NaBH₄ were incorporated into the TiO₂ lattice, disrupting its long-range ordered structure and transforming it into amorphous TiB_xO_y. Simultaneously, the thermal annealing drove the structural reorganization of the support and the re-coordination of Pt atoms. Pt atoms were released from the original Ti-O-Pt coordination environment and preferentially coordinated with the newly generated, near-neutral B atoms within the support, forming Pt-B bonds rather than the Pt-O bonds commonly found in traditional oxide supports. XPS and Extended X-ray Absorption Fine Structure (EXAFS) analyses confirmed that Pt atoms existed as single atoms with a Pt-B coordination environment. This thermal-driven strategy, which accomplishes amorphization, heteroatom doping, and single-atom anchoring simultaneously, enabled the active design of the chemical composition of the amorphous support, thereby tuning the electronic state of Pt to a quasi-zero-valent state and optimizing its hydrogen adsorption free energy.

Electrochemical deposition

Electrochemical deposition constitutes a widely method for preparing amorphous single-atom catalysts. This approach utilizes a pre-synthesized amorphous support as the working electrode, immersed in an electrolyte containing noble metal ions. By precisely controlling the potential, metal ions are selectively reduced and deposited onto the most active sites on the support surface, such as defects, edges, or oxygen vacancies. The advantage of this method lies in its high precision and controllability: by adjusting the deposition potential and time, the loading and dispersion state of single atoms can be precisely controlled. Additionally, the electrochemical environment can induce dynamic reconstruction of the amorphous support surface, thereby constructing unique coordination structures that are difficult to achieve with traditional thermochemical methods. Zhang et al.^[54] employed a cyclic voltammetry-assisted electrochemical deposition method to deposit adjacent Pt single atoms with an interatomic distance of approximately 3.0 Å onto an amorphous FeNiWPB alloy support [Figure 8A]. They first synthesized the amorphous FeNiWPB alloy via a chemical reduction method. Transmission electron microscopy images showed no obvious lattice fringes in the alloy particles, and selected area electron diffraction exhibited diffuse ring patterns [Figure 8B]. Subsequently, this alloy was used as the working electrode, and cyclic voltammetry scans were applied in a solution containing a Pt precursor. During this process, Pt atoms were precisely reduced and deposited onto defect sites on the alloy surface [Figure 8C]. Notably, the introduction of tungsten not only significantly enhanced the structural stability of the catalyst over a wide pH range by forming stable M-W bonds (M = Fe, Pt, Ni) but also induced an electron-rich state at the Pt sites via charge transfer from W to Pt. This regulation optimized the hydrogen adsorption free energy of adjacent Pt atoms from -0.25 eV for PtASSA@FeNiPB to 0.05 eV, closer to the thermoneutral state [Figure 8D]. This W-mediated electronic regulation, combined with the metal-metal synergy between adjacent Pt atoms, collectively endowed the catalyst with excellent hydrogen evolution activity and exceptional stability for up to 600 h in both acidic and alkaline media. Liu et al.^[52] successfully constructed a Pt single-atom catalyst on amorphous Ni(OH)₂ nanosheets using an electrochemical reduction strategy. They first grew amorphous Ni(OH)₂ nanosheet arrays rich in oxygen vacancies on carbon paper via electrochemical transformation. Subsequently, this electrode was subjected to constant potential electrolysis at a specific reduction potential in an alkaline electrolyte containing PtCl₆^2- [Figure 8E]. Due to the presence of numerous coordinatively unsaturated Ni sites on the surface of the amorphous Ni(OH)₂, the reduced Pt ions were not coordinated with oxygen atoms but were instead captured by these exposed Ni atoms, forming unique Pt-Ni metal bonds. Combined results from XAFS and AC-TEM verified that Pt exists in the form of single atoms and forms coordination bonds with Ni sites. This unique coordination environment rendered the Pt single atoms in a negatively charged state, significantly optimizing their adsorption capacity for hydrogen intermediates and promoting the hydrogen spillover process between the amorphous Ni(OH)₂ and Pt sites, thereby substantially enhancing their hydrogen evolution performance under alkaline conditions.

Figure 8. (A) Schematic diagram of the synthesis process; (B) The TEM image and (C) HAADF-STEM image of Pt_ASSA@FeNiWPB; (D) Free energy evolution of Pt_ASSA@FeNiWPB and Pt_ASSA@FeNiPB during the processes of dual H₂O dissociation and two H^* adsorption steps. Figure 8A-D is reprinted with permission from^[54]. Copyright 2026, John Wiley and Sons; (E) Schematic illustration of Pt-SA/a-Ni(OH)₂ growth. Reprinted from^[52], under CC BY 4.0 license. ASSA: Atomic-spaced single-atom; TEM: transmission electron microscopy; HAADF-STEM: high-angle annular dark-field scanning transmission electron microscopy; SA: single atom.

Electrocatalytic hydrogen evolution reaction

The electrocatalytic HER, as the core cathodic reaction in water electrolysis for hydrogen production, exhibits catalytic efficiency that is critically governed by the catalyst’s atomic arrangement, electronic structure, and active site characteristics. Amorphous catalytic materials, defined by long-range disorder and short-range order, overcome the lattice constraints of traditional crystalline materials, thereby emerging as a key material system for enhancing both the intrinsic activity and stability of HER. Single-phase amorphous electrocatalytic materials represent the fundamental research system in the HER field. Their core advantages originate from the intrinsic structure and electronic properties conferred by the disordered atomic arrangement^[110,111]. Amorphous materials lack long-range crystalline order, and their surfaces are abundant in dangling bonds, coordinatively unsaturated sites, and structural defects. Compared to the regular crystal facets of crystalline materials, these sites are more prone to serve as active centers for H adsorption and activation. Concurrently, the flexible atomic network can adaptively adjust the H adsorption configuration, overcoming the inherent bottleneck of crystalline materials in balancing adsorption and desorption. Taking amorphous PtNiP nanoparticles prepared by the flash Joule heating method^[112] as an example, their disordered atomic arrangement reduces the average coordination number of Pt/Ni to 11.5, lower than the coordination number of 12 in crystalline close-packed structures, exposing a large number of low-coordination metal sites [Figure 9A]. This material achieves an overpotential of only 14 mV at a current density of 10 mA cm^-2 in acidic electrolyte, with a Tafel slope as low as 18 mV dec^-1, and a mass activity five times higher than that of commercial Pt/C [Figure 9B]. Furthermore, the isotropic structure of amorphous materials can alleviate lattice stress during the HER process, inhibit aggregation of active sites and structural collapse, demonstrating superior cycling stability compared to crystalline materials.

Figure 9. (A) Coordination number (CN) distribution of Ni/Pt and P within amorphous PtNiP. The average CN values for Ni/Pt and P are determined to be 11.5 and 8.7, respectively; (B) Polarization curves comparing PtNiP ANPs, PtNi CNPs, and commercial Pt/C recorded at the first cycle and after the 10,000th CV cycle. Figure 9A and B is reprinted with permission^[112]. Copyright 2025, American Chemical Society. (C) Nyquist plots of the electrodes; (D) HER performance on pristine NiCo₂O₄ (311) and defective NiCo₂O₄ (311)-SV_O1; (E) LSV curves recorded for Co₃O₄-CC, NiCo₂O₄-CC, and NiCo₂O₄-B-CC for HER in 1 m KOH; (F) Tafel plots of Co₃O₄-CC, NiCo₂O₄-CC and NiCo₂O₄-B-CC toward HER in 1 m KOH. Figure 9C-F is reprinted with permission^[49]. Copyright 2024, John Wiley and Sons; (G) First-shell fitting result derived from the Pt L₃-edge EXAFS spectra and the corresponding Re[k²χ(k)] oscillations of Pt/TiB_xO_y; (H) Mass activity of Pt/TiB_xO_y and Pt/C evaluated at various potentials. Figure 9G and H is reprinted with permission^[51]. Copyright 2022, American Chemical Society; (I) Binding energie variations for Pt/a-Ni(OH)₂ catalysts as a function of electrodeposition time. Reprinted from^[52], under CC BY 4.0 license. RHE: Reversible hydrogen electrode; ANP: amorphous nanoparticle; CNP: crystalline nanoparticles; CC: carbon cloth; HER: hydrogen evolution reaction; LSV: linear sweep voltammetry; EXAFS: extended X-ray absorption fine structure.

Single amorphous systems suffer from issues such as insufficient intrinsic conductivity, uneven distribution of active sites, and limited thermodynamic stability of the structure, restricting their application under high current densities. To overcome the performance bottlenecks of pure amorphous materials, amorphous/crystalline heterojunction emerged. Through strong coupling between the crystalline and amorphous phases, they achieve a “1 + 1 > 2” synergistic effect, becoming a mainstream design direction for HER electrocatalysts. This system perfectly integrates the high conductivity and structural stability of the crystalline phase with the high-density active sites and flexible electronic structure of the amorphous phase. Charge redistribution, lattice distortion, and local electronic state regulation at the interface between the two phases further optimize reaction kinetics and intermediate adsorption behavior. A typical representative is the spinel-type NiCo₂O₄ amorphous/crystalline heterojunction constructed via NaBH₄ reduction^[49]. This system precisely regulates the proportion of crystalline phases through mild reduction treatment, simultaneously introducing high-density oxygen vacancies and heterointerfaces, achieving synergistic optimization of electronic structure, active sites, and charge transport capability. In this heterojunction, the crystalline NiCo₂O₄ framework provides continuous electron transport channels, significantly enhancing the intrinsic conductivity of the material [Figure 9C], addressing the issue of slow carrier mobility in the amorphous phase. The amorphous surface layer retains the characteristics of a long-range disordered structure, generating a large number of coordinatively unsaturated sites and oxygen vacancies, serving as efficient active centers for H₂O dissociation and H adsorption, while also enhancing the material's corrosion resistance and catalytic stability. Significant interface charge redistribution occurs between the amorphous and crystalline phases, driving electrons transfer from the amorphous to the crystalline phase. his modulation optimizes the d-band center of active metal sites, tunes the hydrogen adsorption-desorption energy toward thermodynamically neutrality, and significantly lowers the energy barrier of the HER rate-determining step [Figure 9D]. This amorphous/crystalline heterojunction exhibits excellent HER performance in alkaline electrolyte, with an overpotential of only 26 mV at 10 mA cm^-2 [Figure 9E], a Tafel slope as low as 106 mV dec^-1 [Figure 9F], and stable operation for 20 h at 10 mA cm^-2, far outperforming the pure crystalline NiCo₂O₄ catalyst.

Driven by the pursuit of ultimate atomic utilization and enhanced intrinsic activity in electrocatalysis, amorphous-supported single-atom catalysts have emerged as a frontier hotspot research focus for the HER. Leveraging the distinctive coordination environment of amorphous supports and the maximal atomic utilization of single atoms, a synergistic “amorphous substrate-single atom” architecture is constructed, surpassing the performance limitations of traditional catalysts. The disordered structure, abundant defects, and dangling bonds of amorphous supports provide numerous anchoring sites for single atoms, inhibiting their agglomeration. Simultaneously, the electronic state and coordination structure of single atoms can be precisely regulated through the coordination environment, endowing single atoms with special electronic properties unattainable with traditional crystalline supports. Taking the amorphous TiB_xO_y-supported Pt single atoms^[51] as an example, the near-neutral B atoms in the amorphous support preferentially coordinate with Pt, avoiding Pt oxidation caused by Pt-O bonds [Figure 9G]. Through charge transfer balance between B-O and B-Pt, the nucleophilicity of Pt is regulated to an optimal state, achieving an ideal H adsorption energy, with a mass activity of 37.8 A mg^-1 Pt in acidic HER, 34 times higher than that of commercial Pt/C [Figure 9H]. In the amorphous Ni(OH)₂ supported negatively charged Pt single-atom system^[52], the oxygen vacancies and low-coordination Ni sites of the amorphous substrate anchor Pt and form Pt-Ni bonds, rendering Pt in a negatively charged state (Pt^δ⁻) [Figure 9I]. The amorphous Ni(OH) ₂ efficiently dissociates H₂O and transfers hydrogen to the Pt sites via hydrogen spillover, substantially increasing the hydrogen coverage on the Pt surface. In alkaline HER, this catalyst achieves an overpotential of only 48 mV at a current density of 1000 mA cm^-2 mg^-1, far exceeding the performance of commercial Pt/C.

Electrocatalytic oxygen evolution reaction

Similar to the HER, amorphous materials also exhibit significant structural and performance advantages in the electrocatalytic OER. As a four-electron transfer anodic reaction, the OER has slow kinetics and a higher energy barrier, imposing more stringent requirements on the active sites, conductivity, and structural stability of the catalyst. Single-phase amorphous oxide materials, characterized by long-range structure disordered, are rich in oxygen vacancies, coordinatively unsaturated sites, and tunable metal valence states. They can effectively optimize the adsorption energies of intermediates such as *OH, *O, and *OOH, promote surface reconstruction, and enhance intrinsic activity, generally outperforming their corresponding crystalline catalysts. In amorphous cobalt oxide (CoO_x) materials^[113], the disordered atomic arrangement breaks the regular lattice restrictions of crystalline Co₃O₄, exposing more mixed-valence Co²⁺/Co³⁺ active sites on the surface. Meanwhile, the reducing effect of VS₂ can induce the formation of an amorphous phase with coexisting multiple species, including metallic Co, CoO, Co(OH)₂, and Co₃O₄, significantly increasing the active site density and charge transport capability. In 0.1 M KOH electrolyte, the amorphous CoO_x/VS₂/CP exhibits a significantly lower overpotential at 10 mA cm^-2 and a smaller charge transfer resistance compared to crystalline Co₃O₄/CP [Figure 10A], demonstrating faster adsorption/desorption kinetics for OER intermediates, and a higher current density retention after 100 cycles [Figure 10B], with overall stability comprehensively superior to the crystalline control sample.

Figure 10. (A) Nyquist plots of EIS spectra recorded for CoO_x/VS₂/CP (0.5C), Co₃O₄/CP (0.5C), VS₂/CP, and CP at an applied potential of 1.6 V vs. RHE with a voltage amplitude of 10 mV over a frequency range of 0.1 Hz to 100 kHz; (B) Current densities of CoO_x/VS₂/CP samples with varying Co deposition amounts, measured at 1.6 V during the 1st and 100th cycles. Data were acquired via CV-averaged linear sweep voltammetry over 100 cycles within a potential range of 1.23 to 1.83 V vs. RHE. Reprinted from^[113], under CC BY 4.0 license; (C) Schematic diagram of the OER mechanism governed by synergistic coupling at the crystalline/amorphous interface; (D) Corrosion current densities and corresponding potentials for three catalysts in natural seawater. Figure 10C and D is reprinted with permission^[114]. Copyright 2026, Elsevier; (E) LSV measurements with in-situ iR compensation were performed for a-NMO, a-RNMO, c-NMO and RuO₂ catalysts; (F) Tafel curves fitted from the LSV data in panel (E); (G) Polarization curve recorded for the electrolyzer operated without iR-compensation, where a-RNMO served as the anode and Pt/C as the cathode, in flowing 1 M KOH at 80 °C. Inset: schematic of AEMWE cell. Figure 10e-g is reprinted with permission^[53]. Copyright 2025, Springer Nature. SCE: Saturated calomel electrode; FcNi-BDC: two-dimensional ferrocene-based metal-organic frameworks; a-RNMO: Ru single atoms onto amorphous NiMoO_x; c-NMO: crystalline NiMoO₄; RNMO: rare-earth nickel mixed oxide; RHE: reversible hydrogen electrode; MEA: membrane electrode assembly; EIS: electrochemical impedance spectroscopy; CP: carbon paper; OER: oxygen evolution reaction; LSV: linear sweep voltammetry; iR: iR compensation; CV: cyclic voltammetry; AEMWE: anion exchange membrane water electrolyzer.

Amorphous/crystalline heterojunction achieve synergistic enhancement through interfacial coupling between the two phases. The crystalline component provides efficient electron transport pathways, while the amorphous component offers high-density active sites. Interfacial charge redistribution further modulates the d-band center, significantly reducing the OER overpotential and enhancing cycle life, demonstrating outstanding performance in both alkaline and seawater systems. Taking the heterojunction constructed from crystalline ferrocene-based MOF [FcNi-terephthalic acid (FcNi-BDC)] and amorphous NiOOH^[114] as an example, the crystalline MOF framework provides continuous electron transport channels and structural support, enhancing material conductivity and mechanical stability. The amorphous NiOOH retains a large number of coordinatively unsaturated sites and oxygen vacancies, providing ample OER active centers. A Type-II heterojunction and built-in electric field are formed at the interface between the two phases, driving directional charge transfer, increasing the proportion of Ni³⁺ active species, and optimizing the adsorption free energy of *OOH intermediates [Figure 10C]. In an alkaline seawater system, this heterojunction catalyst achieves an overpotential of only 329 mV at an industrial current density of 1000 mA cm^-2, operates stably for 400 h at 500 mA cm^-2, and has a corrosion current density as low as 26 μA cm^-2 [Figure 10D], exhibiting excellent Cl^- corrosion resistance, making it ideally suited for the harsh conditions of seawater electrolysis.

Building on this foundation, amorphous-supported single-atom catalysts have emerged as a cutting-edge direction in the OER field. The abundant defects and flexible coordination environment of amorphous materials enable efficient anchoring and electronic modulation of single atoms, thereby achieving ultra-high intrinsic activity and atomic utilization at extremely low noble metal loadings. This configuration simultaneously lowers the reaction energy barriers while maintaining excellent stability, providing a novel design concept for high-efficiency, low-cost, and long-life OER electrocatalysis. Taking the amorphous NiMoO_x-anchored Ru single-atom catalyst (a-RNMO)^[53] as an example, the numerous dangling bonds, oxygen vacancies, and disordered coordination environment of the amorphous NiMoO_x support enable highly dispersed and uniform anchoring of Ru single atoms, preventing agglomeration and deactivation, with a Ru loading of only 2.28 at.%. In 1 M KOH electrolyte, this catalyst achieves a low overpotential of 220 mV at a current density of 10 mA cm^-2 [Figure 10E], a Tafel slope of only 45.15 mV dec^-1 [Figure 10F], and a mass activity over 140 times higher than that of commercial RuO₂. In an anion exchange membrane water electrolyzer (AEMWE), it req G uires only 1.78 V cell voltage at a high industrial current density of 1 A cm^-2 [Figure 10] and operates stably for 100 h without significant decay. Through the synergistic electronic regulation of the amorphous structure and single atoms, a combined breakthrough in activity, stability, and economy has been achieved.

The structural advantages of amorphous materials - including abundant coordinatively unsaturated sites, flexible atomic coordination environments, and tunable electronic structures - extend well beyond electrocatalysis into photocatalysis. Specifically, for CO₂ photoreduction, a multi-electron process heavily reliant on efficient charge separation and intermediate activation, these features offer a fundamentally new design paradigm.