High-Entropy Alloy Catalysts toward Multi-Functionality: Synthesis, Application, and Material Discovery

Abstract

High-entropy alloy (HEA) catalysts have attracted tremendous research interest owing to their versatile performances in various applications. However, research on HEA catalysts remains in the early stages of exploration, and the fabrication process, element selection, and application remain difficult. Herein, we summarize the current literature on HEA catalysts from the viewpoint of facile synthesis routes, tunable morphologies, attractive applications, and material discoveries related to machine learning and high-throughput experiments. Finally, perspectives and concepts are presented to design the desired multifunctionality HEA catalysts.

1. Introduction

High-entropy alloys (HEAs) are solid solutions composed of at least five metallic elements with atomic concentrations ranging from 5% to 35%.^1–5) Their high configurational entropy, which allows the existence of a stable solid–solution phase, is beneficial for enhancing strength and ductility. The sluggish diffusion effect can be attributed to the varying potential energies of various elemental atoms compared with those of conventional alloys. Sluggish diffusion is important for phase/elemental stability at elevated temperatures because lowering the atomic diffusion on the surface helps maintain HEA nanostructures. In addition, the crystal lattice can face severe distortions, facilitating the formation of defects and strains. These defects and strains can significantly enhance the catalytic activity of HEAs, making HEAs excellent candidates for catalytic applications. The cocktail effect is a figurative expression that represents the synergistic response from the various elemental components in HEAs. It is not considered a superposition of the characteristics of the individual elements. This effect shows that HEAs can be utilized as a functional material based on the synergetic effect among different elements. Naturally, their application as catalysts has recently gained attention and research has progressed very rapidly. As the active surfaces of HEAs consist of millions of possible atomic arrangements of various elements, optimal properties may be realized at some surface sites to overcome the transitional catalysts.

In response to the recent increasing interest, rapid progress, and numerous trials in this field, we aim to feature the important advances in HEAs catalysts and other functional applications. In particular, we focus on recent achievements describing the intriguing functionalities of high-entropy catalysts as green materials. We summarize the facile synthesis route, tunable morphology with a high surface area, versatile applications, such as water splitting and related energy conversion, and data-driven material discovery to determine the optimal composition from the vast element choices. In the last section, we present our perspectives on the challenges and future research directions to the best of our knowledge.

2. Methodology of Synthesis

We introduce several facile methodologies for the synthesis of HEA catalysts/nanostructures.

2.1 Mechanical alloying

Mechanical alloying is a straightforward method that uses mechanical force to obtain HEA powders with five or more elements. Figure 1(a) shows a schematic of the mechanical alloying process used to synthesize HEAs.⁶⁾ The desired pure metal powders are mixed proportionally and placed in a high-energy planetary ball mill with a milling medium, such as steel balls. Under the strong collision and agitation of steel balls, different metal powders undergo repeated cold-welding, milling, and rewelding processes.⁷⁾ Mechanical alloying is superior to rapid solidification processing as a non-equilibrium processing tool for realizing HEAs with homogeneous structures and compositions, even though these elements are immiscible systems from a thermodynamic point of view. For example, Lv et al. synthesized an AlCoCrTiZn HEA powder by mechanical alloying, which exhibited significant efficiency in degrading azo dyes.⁸⁾ The particles in this HEA powder have a size range of 0.5–10.0 µm and highly irregular shape, and their surfaces were rough and contained several defects. HEA nanoparticle (NPs) AuAgPtPdCu prepared using a cast cum cryomilling process in an Ar environment had notable average small size of ∼16 nm with homogeneously distributed elements in a single phase, which demonstrated the efficient CO₂ electroreduction capability.⁹⁾

Fig. 1

(a) Schematic of the mechanical alloying process for the synthesis of HEAs. Reprinted with permission from Ref. 6). Copyright 2020 American Chemical Society. (b) Wet-chemistry method for the precise and controllable synthesis of HEA NPs. Reprinted from Ref. 10) (CC BY 4.0). (c) Dealloying fabrication process. Reprinted with permission from Ref. 15). Copyright 2021 American Chemical Society. (d) Schematic of the roll-to-roll process for the synthesis using the microwave heating method. Reprinted with permission from Ref. 18). Copyright 2021 American Chemical Society. (e) Schematic of the preparation of HEA in molten CaCl₂. Reprinted with permission from Ref. 20). Copyright 2022 American Chemical Society. (f) Schematic of the fast-moving bed pyrolysis experimental setup for synthesizing HEA NPs. Reprinted from Ref. 21) (CC BY 4.0). (g) Schematic of the femtosecond-laser direct writing method. Reprinted with permission from Ref. 22). Copyright 2022 Wiley-VCH GmbH. (h) Illustration of the progress of sample temperature during electrical current passage. Reprinted from Ref. 25) (CC BY 4.0).

2.2 Wet-chemical approach

The wet-chemical approach is the most common method for preparing HEA NPs and nanocrystals. Wet-chemical synthesis involves chemical reactions in the solution phase using precursors under appropriate experimental conditions without requiring complicated equipment. Figure 1(b) shows a schematic of the wet chemistry for the synthesis of HEAs nanocrystals.¹⁰⁾ The solvothermal/hydrothermal approach is a simple and scalable method widely used to prepare inorganic nanomaterials. In a typical solvothermal process, water or organic solvents act as the reaction media in a closed vessel. The reaction temperature was higher than the boiling point of the solvent. As a result, the solvent was autogenerated at a high pressure, and the crystallinity of the as-synthesized nanocrystals was improved. Hot injection is another wet-chemical synthesis method for preparing monodispersed colloidal nanocrystals with uniform shape and size. This method involves the rapid injection of highly reactive reactants into a surfactant in a hot solution. HEA NPs synthesized via wet-chemistry methods showed excellent tunable sizes and uniform dispersion. Zhang et al. synthesized RuFeCoNiCu HEA NPs prepared in the oil phase at low temperatures (≤250°C) and atmospheric pressure, and applied them to the electrocatalytic nitrogen reduction reaction.¹¹⁾ Wu et al. obtained HEA NPs of all six platinum group metals, Ru, Rh, Pd, Os, Ir, and Pt, by the hot-injection method.¹²⁾ Minamihara et al. developed ultrasmall IrPdPtRhRu HEA NPs (1.32 ± 0.41 nm) using a continuous-flow reactor with a liquid-phase reduction method.¹³⁾ The average size obtained in this work is the smallest value reported for an HEA. This flow synthesis can also provide high productivity with a high reproducibility of at least 500 g/day, which is higher than that of the other synthesis methods.

2.3 Dealloying

Dealloying refers to the selective leaching of one or more components from a solid solution alloy or compound to produce a residual nanoporous structure.¹⁴⁾ Dealloying is recognized as a facile method for fabricating bicontinuous nanoporous metals of tunable sizes. Figure 1(c) shows an example of the dealloying process for synthesizing highly porous HEA catalysts.¹⁵⁾ Precursor alloys with nominal compositions of Al_95.9Co₁Fe₁Mo₁Cr₁Pt_0.1 (at%) were prepared by melting pure melts using an arc-melting furnace, and the alloy ribbons were fabricated by melt spinning. Simple dealloying in a NaOH solution efficiently realized a hierarchical nanoporous HEA. Yu et al. designed single-phase Mn-based precursor alloys Mn₇₀Ni_7.5Cu_7.5Co_4.2V_4.2Fe₂Mo₂Pd_0.5Pt_0.5Au_0.5Ru_0.5Ir_0.5 containing 12 components. They fabricated free-standing nanoporous HEAs containing uniformly distributed 12 metal elements by one-step dealloying in a 1.0 M (NH₄)₂SO₄ solution.¹⁶⁾ Joo et al. utilized liquid-metal dealloying, a unique technique for fabricating non-noble porous materials used in structural, functional, and medical products to prevent oxidation in a metallic melt to develop nanoporous TiVNbMoTa HEAs.¹⁷⁾ The smallest HEA synthesized at 873 K was approximately 10 nm, which is one order of magnitude smaller than those of conventional porous materials. The exceptional stability of this material was achieved through the implementation of slow surface diffusion obtained using a high-entropy design.

2.4 Microwave heating

Rapid heating methods, which are different from conventional furnaces, are being developed to realize the synthesis of nonequilibrium materials. Qiao et al. demonstrated an efficient and scalable microwave heating method using carbon-based materials as substrates to fabricate HEA NPs with a uniform size, as shown in Fig. 1(d).¹⁸⁾ Because of the abundant functional group defects that can efficiently absorb microwaves, reduced graphene oxide is employed as a model substrate to produce an average temperature as high as ∼1850 K within seconds to synthesize PtPdFeCoNi HEA NPs with a size of ∼12 nm and uniform elemental mixing. This microwave radiation-induced Joule heating method has high potential for use in the roll-to-roll process for the scalable production of HEA nanomaterials.

2.5 Molten salt solid-state reduction

The solid-state reduction of molten salts is an alternative strategy for the direct electrochemical transformation of solid oxides into HEAs in high-temperature molten salts. The alloy composition can be tailored by mixing the oxide precursors in a certain stoichiometric ratio.¹⁹⁾ Li et al. fabricated a homogeneous face-centered cubic FeCoNiMnMo HEA by electrolyzing a solid oxide mixture in a molten salt of CaCl₂ at 900°C, as shown in Fig. 1(e).²⁰⁾ The mixed oxide powder was reduced to a coral-like powder with a diameter of 1–2 µm. Concomitant solid-state electrometallization and alloying processes prevent phase segregation. The as-prepared FeCoNiMnMo showed good electrocatalytic activity for the oxygen evolution reaction (OER) and excellent stability for over 1000 h without significant decay.

2.6 Fast-moving bed pyrolysis

Wet impregnation followed by reductive pyrolysis in programmed temperature heating is the most popular method for preparing NPs supported on granular supports. However, the stabilization of HEAs on granular supports, such as active carbon, alumina oxide, and zeolite, by conventional Joule heating remains a difficult challenge. Therefore, Gao et al. developed a fast-moving bed pyrolysis strategy, as shown in Fig. 1(f), following wet impregnation to prepare ultra-small and highly dispersed HEA NPs with up to 10 elements, MnCoNiCuRhPdSnIrPtAu.²¹⁾ The representative quinary (FeCoPdIrPt) HEA NPs exhibited high activity and exceptional stability toward hydrogen evolution during water splitting.

2.7 Pulse laser

To date, the aforementioned synthesis methods require rigorous conditions, including high pressure, high temperature, and inert atmospheric protection. To overcome this problem, Hegde et al. demonstrated a novel method for the in situ synthesis and coating of HEAs and multi-metal oxide nano–microparticles via a one-step femtosecond-laser direct writing (FsLDW) process, as shown in Fig. 1(g).²²⁾ The FsLDW was used to fabricate binder-free electrodes at room temperature in an open atmosphere. In-depth investigations of the ink formulation and laser parameters were performed to identify the optimal parameters for a successful LDW. Similarly, Wang et al. applied a laser-scanning ablation approach using a pulsed fiber nanosecond laser to synthesize a library of HEA NPs at atmospheric temperature and pressure.²³⁾ The laser ablates the corresponding NP precursors in alkanes, thereby enabling the formation of HEAs and high-entropy compounds, such as oxides, sulfides, phosphides, nitrides, and borides within a 5-ns pulse.

2.8 High-temperature carbothermal shock

High-temperature carbothermal shock (CTS) is the first general method for controllable and efficient synthesis of HEA NPs with a single-phase structure and uniform dispersion, as demonstrated by Yao et al.²⁴⁾ The CTS method employs electrically shocked mixed-metal-salt precursors dispersed on conductive carbon supports, resulting in flash heating and cooling of the metal precursors on the carbon substrate to provide HEA NPs. The temperature reached ∼2000 K, shock duration was 55 ms, and ramp rate was on the order of 10⁵ K/s. The maximum temperature of the CTS method was sufficiently high to decompose any metal salt; therefore, uniform mixing of nearly any metallic combination is realized. The precise control of the heating parameters can effectively tune the particle size and distribution. PtPdRhRuCe NPs were found to be effective catalysts for ammonia oxidation; thus, this synthetic approach has the potential to yield new catalysts. Abdelhafiz et al. also demonstrated a ternary-to-sernary (FeNiCoCrMnV) HEA using CTS synthesis.²⁵⁾ The electronic current passing through the gas diffusion layer increased its temperature to as high as 1300–1700°C, as shown in Fig. 1(h).

3. Tunable Morphology

NPs are the major morphology of HEA catalysts; however, various other morphologies have also been reported. Therefore, we introduced a wide range of shapes, from nanoclusters to bicontinuous nanoporous forms, as extended nanostructure shapes, as summarized in Fig. 2.

Fig. 2

(a) Template synthesis of multimetallic sub-nanoclusters, whereby metal elements were precisely accumulated on DPA. Reprinted from Ref. 26) (CC BY 4.0). (b) Platinum-group-metal HEA NPs. Reprinted with permission from Ref. 12). Copyright 2020 American Chemical Society. (c) 3D models and enlarged atomic model of HEA subnanometer nanowires. Reprinted from Ref. 28) (CC BY 4.0). (d) 2D HEA nanosheet synthesis by tuning the hydrogen binding energy. Reprinted with permission from Ref. 30). Copyright 2022 American Chemical Society. (e) SEM pictures of the surface of mesoporous HEA-based particles. Reprinted with permission from Ref. 31). Copyright 2022 American Chemical Society. (f) HAADF–STEM image and HAADF–EDS elemental maps of an individual octonary CrMnFeCoNiPdRuIr hollow HEA NPs. Scale bar: 100 nm. Reprinted with permission from Ref. 32). Copyright 2020 Wiley-VCH GmbH. (g) Schematic of nanoporous ultra-HEAs containing up to 14 elements. Reprinted from Ref. 34) (CC BY 4.0). (h) Schematic of HEA aerogels with boosted HCOOH generation. Reprinted with permission from Ref. 37). Copyright 2022 Wiley-VCH GmbH.

3.1 Nanocluster

Tsukamoto et al. demonstrated the template synthesis of multimetallic sub-nanoclusters of sub 1 nm size using a phenylazomethine dendrimer as a macromolecular template,²⁶⁾ as shown in Fig. 2(a). A dendrimer with phenylazomethine moieties (DPA) was used as a template for the synthesis of monodispersed sub-nanoclusters. Precise accumulation of various metal salts on the template was accomplished using this template strategy. For example, multimetallic sub-nanoclusters composed of five elements (Ga₁In₁Au₃Bi₂Sn₆) were synthesized. This synthesis method provides the capability to mix different elements in various combinations to synthesize new catalysts at the subnanometer scale.

3.2 NPs

NPs are the most popular shapes of catalysts, as shown in Fig. 2(b).¹²⁾ Naturally, HEA NPs have received considerable attention in recent years because of their multi-elemental composition and homogeneously mixed solid-solution state. NPs can provide not only an enormous number of combinations for material discovery but also excellent catalytic properties. Typically, in a thermal shock process (Fig. 1(h)), a solid-solution state is achieved with the rapid heating of the precursors to a high temperature, and the short heating duration and subsequent rapid quenching can retain a uniform structure and small particle size. Owing to the nature of their nanosize and tunable electronic states, the performance of high-entropy NPs has been recognized in a wide range of catalysis and energy technologies.²⁷⁾

3.3 Nanowire and nanoribbon

Zhan et al. demonstrated unique PtRuNiCoFeMo HEA sub-nanometer nanowires via a wet-chemical approach as highly active and durable catalysts for the alkaline hydrogen oxidation reaction, as shown in Fig. 2(c).²⁸⁾ The mass and specific activities of these products reach 6.75 A mg_Pt+Ru⁻¹ and 8.96 mA cm⁻², which are higher than those of commercial PtRu/C and Pt/C, respectively. Density functional theory (DFT) calculations confirmed that the strong interactions between different metals regulate their electronic structures, thereby enhancing the catalytic activity. Tao et al. demonstrated a general synthetic method to fabricate superthin PtPdIrRuAg HEA subnanometer ribbons with a layer thickness of 0.8 nm via the galvanic exchange reaction between different metal precursors with Ag nanowire template.²⁹⁾ The representative quinary HEA–PtPdIrRuAg nanoribbon is an efficient and stable electrocatalyst for oxygen reduction reaction (ORR) in alkaline electrolytes.

3.4 Nanosheet

Fu et al. reported the discovery of a Pt-free combination, PdMoGaInNi, via computer-facilitated screening to optimize electrocatalysts for hydrogen evolution reaction (HER). They then synthesized a two-dimensional (2D) HEA nanosheet via a wet-chemical approach by mixing metal precursors in oleylamine using excess Mo(CO)₆ as the reducing agent and generated CO in situ as the capping agent,³⁰⁾ as shown in Fig. 2(d). The PdMoGaInNi nanosheet exhibited excellent stability and outperformed the HER activity of Pt/C and Pd/C, offering a promising replacement for Pt-based electrocatalysts in water electrolyzers.

3.5 Macroporous and mesoporous forms

A hierarchical architecture is a key requirement for various functional purposes. De Macro et al. introduced ordered mesoporous and hierarchical meso/macroporous HEA-based nanocrystals using soft-templating approaches,³¹⁾ as shown in Fig. 2(e). The formation mechanism is as follows: The metal chloride precursors were dissolved in deionized water and mixed with an aqueous suspension of polymer latex. The spray–drying process generates an aerosol of droplets containing latex particles and dissolved metal chloride salts by solution atomization. After water evaporation, the droplets became hybrid spheroidal microparticles composed of organic latex beads and an inorganic metal chloride salt matrix. Upon thermal annealing, the polymer latex template is decomposed, and meso/macroporous nanostructured particles remained. The pore size can be tuned by adjusting the diameter of the polymer template beads. This approach does not require extremely high process temperatures or flammable or toxic solvents, unlike other common methods for the synthesis of HEAs.

3.6 Hollow NPs

Hollow NPs are an important class of nanomaterials with shells that encapsulate large inner voids. This unique hollow structure has a high specific surface area, low density, and reduced path lengths for mass and charge transfer. Wang et al. developed a continuous “droplet-to-particle” method for synthesizing hollow HEA NPs by introducing a gas-blowing agent and transient high-temperature heating with a uniform mixing of up to eight distinct elements.³²⁾ Figure 2(f) shows the high-angle annular dark field (HAADF)-scanning transmission electron microscopy (STEM) image and HAADF–energy dispersive spectrometry (EDS) elemental maps of the individual octonary-CrMnFeCoNiPdRuIr hollow HEA NPs. As an application of Li–O₂ battery operation, hollow RuIrFeCoNi HEA NPs can achieve a record-high current density of 2000 mA gcat.⁻¹ when used as a cathode catalyst, compared to a traditional solid catalyst. No apparent morphological decomposition or degradation was observed for the spent catalyst after 80 reversible operation cycles.

The composition, microstructure, and crystallinity of metal–organic frameworks (MOFs) can be easily tuned by adjusting the metal ions, organic ligands, and synthetic conditions. Hu et al. demonstrated a porous hollow high-entropy MOF-74 consisting of Mn, Fe, Co, Ni, Cu, and Zn using a one-pot hydrothermal method.³³⁾ After annealing the MOFs at 650°C, the derived high-entropy nanocomposite comprising metal particles and a metal oxide (ZnO) exhibited superior ORR performance. The unique hollow structure and synergistic effect of the multicomponent between the formed ZnO and alloy NPs achieved a good ORR performance.

3.7 Nanoporous form

The nanoporous form is the major shape obtained using the dealloying technique, as mentioned in Section 2.3. The bicontinuous form requires no support/substrate and has the benefit of using an electrode compared with the NPs system. Cai et al. demonstrated the facile synthesis of nanoporous ultra-HEAs containing up to 14 elements, AlAgAuCoCuFeIrMoNiPdPtRhRuTi, as shown in Fig. 2(g).³⁴⁾ They exhibit high catalytic activities and electrochemical stabilities in the HER and OER in acidic media, which are superior to those of commercial Pt/graphene and IrO₂ catalysts. Qiu et al. synthesized senary-AlNiCuPtPdAu with ligament sizes of 2–3 nm by dealloying the designed precursor alloys.³⁵⁾ With a naturally formed thin oxide layer of spinel γ-Al₂O₃, the nanoporous AlNiCuPtPdAu HEA exhibited an enhanced structural stability of up to 600°C. Han et al. fabricated a hierarchical porous AlCrFeCoNiW HEA consisting of face-centered cubic and ordered back-centered cubic (BCC) phases by dealloying an Al-enriched BCC phase.³⁶⁾ The porous HEA electrode exhibited excellent HER and OER activities, and outstanding stability under industrial conditions. More recently, Li et al. fabricated PdCuAuAgBiIn HEA aerogels (HEAAs) via a freeze–thaw method, as shown in Fig. 2(h).³⁷⁾ The freeze–thaw method is universal for the preparation of multicomponent HEAAs, such as octonary-PdCuAuAgBiInCoNi HEAAs.

4. Application

4.1 Water splitting

Electrocatalysts are the most widely used HEA catalysts, and the benefits of multiple elements are well recognized. In particular, electrocatalysts for water splitting have been extensively investigated and summarized in several review articles.^38–44) One good example of the “cocktail effect” is demonstrated for the HER by Wu et al., who fabricated an IrPdPtRhRu HEA NPs catalyst and found that this HEA had deeper d-band center locations between Ir and Pt;⁴⁵⁾ thus, it was assumed to have a similar activity as that of the Pt or Ir catalyst. The turnover frequency values of the IrPdPtRhRu HEA NPs in acidic and alkaline solutions were considerably higher than those of the individual elements without the d-band center, as shown in Fig. 3(a). HEAs contain various atomic arrangements with a unique local density of states. OER is another type of water-splitting reaction. Nguyen et al. synthesized a noble-metal-free high-entropy glycerate (HEG) with five metals (Fe, Ni, Co, Cr, and Mn) via a simple solvothermal process.⁴⁶⁾ HEG was found to be more active than its quaternary, ternary, and binary metal glycerate subsystems, as shown in Fig. 3(b). The excellent catalytic activity of HEG is attributed to the synergistic interactions among the metal elements, whereas its excellent electrochemical stability is ascribed to the high-entropy configuration effect.

Fig. 3

(a) Polarization curves of IrPdPtRhRu HEA NPs for HER. Reprinted from Ref. 45) (CC BY 3.0). (b) Polarization linear sweep voltammetry (LSV) curves of unary-metal glycerates for OER. Reprinted from Ref. 46) (CC BY 4.0). (c) Selectivity–activity plots of the CO₂RR/CORR selectivity and CORR activity space. Reprinted with permission from Ref. 48). Copyright 2020 American Chemical Society. (d) Configuration of the Zn–air battery. Reprinted from Ref. 49) (CC BY-NC 3.0). (e) Optical image of an LED light in the daytime powered by two AlNiCoRuMo-based batteries. The inset is the unlit and lit up LED indicator at night. Reprinted with permission from Ref. 50). Copyright 2020 American Chemical Society. (f) Operando XRD results obtained during the first full lithiation/delithiation cycle. The black lines with arrows indicate the as-prepared, fully lithiated, and fully delithiated states as a function of time, with the corresponding potential curve shown on the left side. Reprinted from Ref. 52) (CC BY 4.0).

4.2 Electrochemical CO₂ reduction (CO₂RR)

The efficient conversion of CO₂ into carbon fuels through electrocatalytic reduction plays a key role in sustaining global energy demand. Nellaiappan et al. realized a cryomilled nanocrystalline equiatomic AuAgPtPdCu HEA for the efficient electrochemical reduction of CO₂RR.⁴⁷⁾ Although five elements were used, the electrocatalytic activity was mainly associated with the presence of redox-active Cu metal. The full conversion of CO₂ to gaseous products at a low voltage (−0.3 V RHE) cannot be attained from pure metallic Cu. Pedersen et al. theoretically screened two HEA systems, CoCuGaNiZn and AgAuCuPdPt, as promising catalyst candidates for the CO₂RR and CO reduction reaction (CORR) (Fig. 3(c)).⁴⁸⁾ They predicted the CO and hydrogen adsorption energies of all HEA surface sites by combining DFT and supervised machine learning (ML). This demonstrates the ability of the model to predict valid catalyst candidates without prior knowledge of their catalytic properties or the composition of disordered alloys for optimal catalytic performance.

4.3 Zinc–air battery

The high catalytic activity of HEAs in the OER and ORR for water splitting makes them suitable for preparing an air cathode in a zinc–air battery, as shown in Fig. 3(d). Jin et al. fabricated seven-component HEA nanoclusters (PtPdAuAgCuIrRu) supported on a seven-component nanostructured spinel high-entropy oxide (HEO) (AlNiCoFeCrMoTi) using a two-step alloying–dealloying strategy.⁴⁹⁾ When this high-entropy composite was used as a bifunctional cathode in a zinc–air battery, it also exhibited excellent charge–discharge reversibility. The seven-component HEO was highly active for the OER, whereas the noble-metal-based HEA clusters were responsible for the high ORR activity, resulting in a record-small charge–discharge reversibility. Yu et al. also synthesized free-standing multicomponent nanoporous HEAs consisting of 12 different elements, including noble and non-noble metals (MnNiCuCoVFePdPtAuRuIrMo), by dealloying.¹⁶⁾ Because of the cooperation of these elements, the 12-element HEA showed high activity and durability for HER, OER, and ORR, making it suitable for rechargeable zinc–air batteries. Similarly, Jin et al. fabricated HEA nanowires by dealloying AlNiCoRuX, where X = Mo, Cu, V, and Fe, which are trifunctional electrocatalysts for the HER, OER, and ORR, respectively. Figure 3(e) shows the optical image of a light-emitting diode (LED) in a daytime LED powered by two AlNiCoRuMo-based zinc–air batteries.⁵⁰⁾

4.4 Li-ion battery (LIB)

Apart from alloy systems, HEO has received considerable attention as the most promising cathode material for next-generation LIBs.⁵¹⁾ Sakar et al. demonstrated the reversible lithium storage properties of HEOs, the underlying mechanism, and the influence of entropy stabilization on the electrochemical behavior. They found that the stabilization effect of entropy brings significant benefits for the storage capacity retention of HEOs and greatly improves cycling stability.⁵²⁾ Figure 3(f) shows the operando X-ray diffraction (XRD) pattern during the first full lithiation/delithiation cycle of HEO. The reflections from the rock-salt structure disappeared during the first lithiation because of the formation of small crystallites with sizes below the detection threshold of XRD. Lun et al. demonstrated a cation-disordered rock-salt-type HEO for LIB.⁵³⁾ The cathode compound, composed of six different transition metal (TM) species, was approximately 40% larger than the capacity of the conventional compound. The incorporation of a large number of TM species into the lattice can reduce the extent of short-range order and dramatically improve Li transport. Zhang et al. recently proposed a high-entropy doping strategy to fabricate high-Ni and zero-Co layered cathodes with long cycle lives.⁵⁴⁾ They found that the cathode exhibited nearly zero volumetric change over a wide electrochemical window.

5. Material Discovery

5.1 Screening via computation

The binding energy distribution pattern (Fig. 4(a)) is an important concept for understanding the unique features of HEA catalysts.⁵⁵⁾ For a quinary alloy, at least five binding peaks were observed, and their intensity integral was determined by the probability of the binding atom at the surface. Therefore, determining an element combination is important to achieve a narrow peak with a position as close as possible to the optimal binding energy for the reaction. In reality, the multi-element interaction is complex, and predicting a suitable element combination is not straightforward. DFT calculations may provide most of this information, and the experimental composition trends can be predicted using this model. However, it requires considerable computational power and time to model a single-element combination. Fu et al. employed computational tools to screen and identify promising element combinations for the optimal HEA for the HER, and PdMoGaInNi was identified as the optimal combination for a promising Pt-free HER electrocatalyst, as shown in Fig. 4(b).³⁰⁾ Chun et al. demonstrated the combination of ML approach and experimental tests of PtFeCu NPs for ORR, as shown in Fig. 4(c).⁵⁶⁾ Promising candidates were efficiently screened, and theoretically and experimentally validated. By optimizing the PtFeCu alloy ratio, PtFe_lowCu_high exhibited the best catalytic performance among the ternary samples. Thus, first-principles calculations with an ML approach can be a solution for designing nanocatalysts to fill the gap between experiments and simulations.

Fig. 4

(a) Schematic of the binding energy distribution pattern and plateau current depend on the binding energy shift and peak integral. Reprinted from Ref. 55) (CC BY-NC-ND 4.0). (b) Computational prediction of the desirable composition for HER electrocatalyst. A volcano plot is constructed from the experimentally derived exchange current densities and computationally calculated hydrogen binding energy. Reprinted with permission from Ref. 30). Copyright 2022 American Chemical Society. (c) Schematic of ternary alloy configuration search and theoretical predictions and experimental validation for ORR using a neutral network. Reprinted from Ref. 56) (CC BY-NC-ND 4.0). (d) Schematic of the combinatorial and high-throughput synthesis of uniform multimetallic nanoclusters, the scanning droplet cell setup, and patterned samples on the copper substrate. CE, counter electrode; RE, reference electrode; WE, working electrode. Reprinted from Ref. 57) (CC BY-NC-ND 4.0). (e) Schematic of the high-throughput experimental setup for synthesizing alloys with a large composition space. Reprinted with permission from Ref. 60). Copyright 2021 Wiley-VCH GmbH.

5.2 Screening via high-throughput experiments

Yao et al. reported high-throughput synthesis and evaluation of an extensive series of ultrafine and homogeneous HEA. The achieved components were as follows: 1) a flexible compositional design formulation in the precursor solution phase, 2) ultrafast synthesis of thermal shock heating to reach ∼1,650 K at a duration time of ∼500 ms, and 3) high-throughput electrochemical characterization using a scanning droplet cell, as shown in Fig. 4(d).⁵⁷⁾ This approach is remarkably facile and accessible compared with conventional vapor-phase deposition. Uniform particle size and structure can be used for comparative studies of different samples. They discovered that PtPdFeCoNi and PtPdRhNi HEAs are good candidates for the HER. Batcheler et al. combined simulation, ML, data-guided combinatorial synthesis, and high-throughput characterization. They demonstrated models for predicting the ORR activity based on simple descriptors without sufficient information to make predictions.⁵⁸⁾ A comparison of the data from over 1000 ORR activity measurements and ML-guided predictions correctly predicted the activity maxima of the model system, AgIrPdPtRu. Banko et al. applied a combinatorial strategy to acquire large experimental datasets of 5D composition spaces, RuRhPdIrPt.⁵⁹⁾ Advanced simulations and extensive experimental data can be used to estimate the electrocatalytic activity and solid-solution stability trends in the 5D composition space. High-throughput electrocatalytic measurements were performed, and 2052 linear sweep voltammograms were recorded and evaluated. The maximum electrocatalytic activity was obtained for Ru₂₅Rh₁₅Pd₃₁Ir₁₅Pt₁₄.

For the synthesis of bulk samples, not limited to thin films, Zhu et al. demonstrated a fast and high-throughput method for synthesizing bulk HEAs based on a radio frequency inductively coupled plasma technique, as shown in Fig. 4(e).⁶⁰⁾ Typical samples were fabricated within 40 s per alloy, thereby significantly improving the fabrication efficiency and throughput. More specific reviews on DFT and screening are available elsewhere.⁶¹⁾

6. Outlook

Research on HEA catalysts is still in its infancy. The number of applications is still limited, and the applicability for organic-molecular transformations as complex reactions remains untested. On the one hand, poor selectivity of products is expected because of the involvement of various elements. On the other hand, new molecular transformations could be discovered by unexpected atomic arrangements of various elements on the surface. The visualization of the catalytic origin can be depicted as a network, such as the brain network,²⁷⁾ from the cooperative atomic motion of each element. To understand the catalytic origin, ML and DFT have been used to predict the adsorption energies of key reaction intermediates. Using trained-DFT data only, Lu et al. demonstrated that the ML model prediction was in agreement with the measured activity from prior literature.⁶²⁾ As an advanced characterization method to identify the atomic origin in electrochemistry, in situ electrochemical X-ray photoelectron spectroscopy can be a powerful tool for discovering corrosion-resistant HEA electrodes in acid media.⁶³⁾ Nonetheless, the number of studies on high-throughput experiments is limited, and various elemental combinations in HEAs have not yet been explored. By understanding the catalytic origins of each constituent element more accurately, we can design order-made catalysts for various reactions, such as building blocks of colorful elements, as shown in Fig. 5.

Fig. 5

Model of the future direction of HEA catalyst. The catalytic origins are depicted as complex as brain network where the elements are diversely associated with each other. With the assistance from advanced characterization, high-throughput experiments, and data-driven science, the entire figure can be simplified into tailor-made elemental building blocks.

Acknowledgments

This work was financially supported by KAKENHI (Grant No. JP21H00153 and JP21H02037), and a cooperative program (Proposal Nos. 202012–CRKEQ–0214 and 202205–CRKEQ–0051) of the CRDAM–IMR, Tohoku University.

REFERENCES

• 1)   J.-W.  Yeh,  S.-K.  Chen,  S.-J.  Lin,  J.-Y.  Gan,  T.-S.  Chin,  T.-T.  Shun,  C.-H.  Tsau and  S.-Y.  Chang: Adv. Eng. Mater. 6 (2004) 299–303. doi:10.1002/adem.200300567
• 2)   B.  Cantor,  I.T.H.  Chang,  P.  Knight and  A.J.B.  Vincent: Mater. Sci. Eng. A 375–377 (2004) 213–218. doi:10.1016/j.msea.2003.10.257
• 3)   E.P.  George,  D.  Raabe and  R.O.  Ritchie: Nat. Rev. Mater. 4 (2019) 515–534. doi:10.1038/s41578-019-0121-4
• 4)   D.B.  Miracle and  O.N.  Senkov: Acta Mater. 122 (2017) 448–511. doi:10.1016/j.actamat.2016.08.081
• 5)   S.  Gorsse,  D.B.  Miracle and  O.N.  Senkov: Acta Mater. 135 (2017) 177–187. doi:10.1016/j.actamat.2017.06.027
• 6)   Y.  Xin,  S.  Li,  Y.  Qian,  W.  Zhu,  H.  Yuan,  P.  Jiang,  R.  Guo and  L.  Wang: ACS Catal. 10 (2020) 11280–11306. doi:10.1021/acscatal.0c03617
• 7)   B.S.  Murty and  S.  Ranganathan: Int. Mater. Rev. 43 (1998) 101–141. doi:10.1179/imr.1998.43.3.101
• 8)   Z.Y.  Lv,  X.J.  Liu,  B.  Jia,  H.  Wang,  Y.  Wu and  Z.P.  Lu: Sci. Rep. 6 (2016) 34213. doi:10.1038/srep34213
• 9)   S.  Nellaiappan,  N.K.  Katiyar,  R.  Kumar,  A.  Parui,  K.D.  Malviya,  K.G.  Pradeep,  A.K.  Singh,  S.  Sharma,  C.S.  Tiwary and  K.  Biswas: ACS Catal. 10 (2020) 3658–3663. doi:10.1021/acscatal.9b04302
• 10)   L.  Yu,  K.  Zeng,  C.  Li,  X.  Lin,  H.  Liu,  W.  Shi,  H.-J.  Qiu,  Y.  Yuan and  Y.  Yao: Carbon Energy 4 (2022) 731–761. doi:10.1002/cey2.228
• 11)   D.  Zhang,  H.  Zhao,  X.  Wu,  Y.  Deng,  Z.  Wang,  Y.  Han,  H.  Li,  Y.  Shi,  X.  Chen,  S.  Li,  J.  Lai,  B.  Huang and  L.  Wang: Adv. Funct. Mater. 31 (2021) 2006939. doi:10.1002/adfm.202006939
• 12)   D.  Wu,  K.  Kusada,  T.  Yamamoto,  T.  Toriyama,  S.  Matsumura,  S.  Kawaguchi,  Y.  Kubota and  H.  Kitagawa: J. Am. Chem. Soc. 142 (2020) 13833–13838. doi:10.1021/jacs.0c04807
• 13)   H.  Minamihara,  K.  Kusada,  D.  Wu,  T.  Yamamoto,  T.  Toriyama,  S.  Matsumura,  L.S.R.  Kumara,  K.  Ohara,  O.  Sakata,  S.  Kawaguchi,  Y.  Kubota and  H.  Kitagawa: J. Am. Chem. Soc. 144 (2022) 11525–11529. doi:10.1021/jacs.2c02755
• 14)   I.  McCue,  E.  Benn,  B.  Gaskey and  J.  Erlebacher: Annu. Rev. Mater. Res. 46 (2016) 263–286. doi:10.1146/annurev-matsci-070115-031739
• 15)   Z.  Jin,  J.  Lyu,  Y.-L.  Zhao,  H.  Li,  Z.  Chen,  X.  Lin,  G.  Xie,  X.  Liu,  J.-J.  Kai and  H.-J.  Qiu: Chem. Mater. 33 (2021) 1771–1780. doi:10.1021/acs.chemmater.0c04695
• 16)   T.  Yu,  Y.  Zhang,  Y.  Hu,  K.  Hu,  X.  Lin,  G.  Xie,  X.  Liu,  K.M.  Reddy,  Y.  Ito and  H.-J.  Qiu: ACS Mater. Lett. 4 (2022) 181–189. doi:10.1021/acsmaterialslett.1c00762
• 17)   S.-H.  Joo,  J.W.  Bae,  W.-Y.  Park,  Y.  Shimada,  T.  Wada,  H.S.  Kim,  A.  Takeuchi,  T.J.  Konno,  H.  Kato and  I.V.  Okulov: Adv. Mater. 32 (2020) 1906160. doi:10.1002/adma.201906160
• 18)   H.  Qiao,  M.T.  Saray,  X.  Wang,  S.  Xu,  G.  Chen,  Z.  Huang,  C.  Chen,  G.  Zhong,  Q.  Dong,  M.  Hong,  H.  Xie,  R.  Shahbazian-Yassar and  L.  Hu: ACS Nano 15 (2021) 14928–14937. doi:10.1021/acsnano.1c05113
• 19)   P.  Li,  Y.  Yao,  W.  Ouyang,  Z.  Liu,  H.  Yin and  D.  Wang: J. Mater. Sci. Technol. 138 (2023) 29–35. doi:10.1016/j.jmst.2022.08.012
• 20)   P.  Li,  X.  Wan,  J.  Su,  W.  Liu,  Y.  Guo,  H.  Yin and  D.  Wang: ACS Catal. 12 (2022) 11667–11674. doi:10.1021/acscatal.2c02946
• 21)   S.  Gao,  S.  Hao,  Z.  Huang,  Y.  Yuan,  S.  Han,  L.  Lei,  X.  Zhang,  R.  Shahbazian-Yassar and  J.  Lu: Nat. Commun. 11 (2020) 2016. doi:10.1038/s41467-020-15934-1
• 22)   C.  Hegde,  C.H.J.  Lim,  T.H.  Teng,  D.  Liu,  Y.-J.  Kim,  Q.  Yan and  H.  Li: Small 18 (2022) 2203126. doi:10.1002/smll.202203126
• 23)   B.  Wang,  C.  Wang,  X.  Yu,  Y.  Cao,  L.  Gao,  C.  Wu,  Y.  Yao,  Z.  Lin and  Z.  Zou: Nat. Synth. 1 (2022) 138–146. doi:10.1038/s44160-021-00004-1
• 24)   Y.  Yao et al.: Science 359 (2018) 1489–1494. doi:10.1126/science.aan5412
• 25)   A.  Abdelhafiz,  B.  Wang,  A.R.  Harutyunyan and  J.  Li: Adv. Energy Mater. 12 (2022) 2200742. doi:10.1002/aenm.202200742
• 26)   T.  Tsukamoto,  T.  Kambe,  A.  Nakao,  T.  Imaoka and  K.  Yamamoto: Nat. Commun. 9 (2018) 3873. doi:10.1038/s41467-018-06422-8
• 27)   Y.  Yao,  Q.  Dong,  A.  Brozena,  J.  Luo,  J.  Miao,  M.  Chi,  C.  Wang,  I.G.  Kevrekidis,  Z.J.  Ren,  J.  Greeley,  G.  Wang,  A.  Anapolsky and  L.  Hu: Science 376 (2022) eabn3103. doi:10.1126/science.abn3103
• 28)   C.  Zhan,  Y.  Xu,  L.  Bu,  H.  Zhu,  Y.  Feng,  T.  Yang,  Y.  Zhang,  Z.  Yang,  B.  Huang,  Q.  Shao and  X.  Huang: Nat. Commun. 12 (2021) 6261. doi:10.1038/s41467-021-26425-2
• 29)   L.  Tao,  M.  Sun,  Y.  Zhou,  M.  Luo,  F.  Lv,  M.  Li,  Q.  Zhang,  L.  Gu,  B.  Huang and  S.  Guo: J. Am. Chem. Soc. 144 (2022) 10582–10590. doi:10.1021/jacs.2c03544
• 30)   X.  Fu,  J.  Zhang,  S.  Zhan,  F.  Xia,  C.  Wang,  D.  Ma,  Q.  Yue,  J.  Wu and  Y.  Kang: ACS Catal. 12 (2022) 11955–11959. doi:10.1021/acscatal.2c02778
• 31)   M.L.  De Marco,  W.  Baaziz,  S.  Sharna,  F.  Devred,  C.  Poleunis,  A.  Chevillot-Biraud,  S.  Nowak,  R.  Haddad,  M.  Odziomek,  C.  Boissière,  D.P.  Debecker,  O.  Ersen,  J.  Peron and  M.  Faustini: ACS Nano 16 (2022) 15837–15849. doi:10.1021/acsnano.2c05465
• 32)   X.  Wang et al.: Adv. Mater. 32 (2020) 2002853. doi:10.1002/adma.202002853
• 33)   J.  Hu,  L.  Cao,  Z.  Wang,  J.  Liu,  J.  Zhang,  Y.  Cao,  Z.  Lu and  H.  Cheng: Compos. Commun. 27 (2021) 100866. doi:10.1016/j.coco.2021.100866
• 34)   Z.-X.  Cai,  H.  Goou,  Y.  Ito,  T.  Tokunaga,  M.  Miyauchi,  H.  Abe and  T.  Fujita: Chem. Sci. 12 (2021) 11306–11315. doi:10.1039/D1SC01981C
• 35)   H.-J.  Qiu,  G.  Fang,  Y.  Wen,  P.  Liu,  G.  Xie,  X.  Liu and  S.  Sun: J. Mater. Chem. A 7 (2019) 6499–6506. doi:10.1039/C9TA00505F
• 36)   X.  Han,  Q.  Chen,  Q.  Chen,  Q.  Wu,  Z.  Xu,  T.  Zheng,  W.  Li,  D.  Cui,  Z.  Duan,  J.  Zhang,  J.  Li,  H.  Li,  Z.  Wang,  J.  Wang and  Z.  Xia: J. Mater. Chem. A 10 (2022) 11110–11120. doi:10.1039/D2TA01701F
• 37)   H.  Li,  H.  Huang,  Y.  Chen,  F.  Lai,  H.  Fu,  L.  Zhang,  N.  Zhang,  S.  Bai and  T.  Liu: Adv. Mater. 35 (2023) 2209242. doi:10.1002/adma.202209242
• 38)   X.  Xu,  Z.  Shao and  S.P.  Jiang: Energy Technol. 10 (2022) 2200573. doi:10.1002/ente.202200573
• 39)   Q.  Zhang,  S.  Zhang,  Y.  Luo,  Q.  Liu,  J.  Luo,  P.K.  Chu and  X.  Liu: APL Mater. 10 (2022) 070701. doi:10.1063/5.0097479
• 40)   Y.  Sun and  S.  Dai: Sci. Adv. 7 (2021) eabg1600. doi:10.1126/sciadv.abg1600
• 41)   J.  You,  R.  Yao,  W.  Ji,  Y.  Zhao and  Z.  Wang: J. Alloy. Compd. 908 (2022) 164669. doi:10.1016/j.jallcom.2022.164669
• 42)   H.  Liu,  L.  Syama,  L.  Zhang,  C.  Lee,  C.  Liu,  Z.  Dai and  Q.  Yan: SusMat 1 (2021) 482–505. doi:10.1002/sus2.32
• 43)   X.  Huo,  H.  Yu,  B.  Xing,  X.  Zuo and  N.  Zhang: Chem. Rec. 22 (2022) e202200175. doi:10.1002/tcr.202200175
• 44)   A.K.  Ipadeola,  A.K.  Lebechi,  L.  Gaolatlhe,  A.B.  Haruna,  M.  Chitt,  K.  Eid,  A.M.  Abdullah and  K.I.  Ozoemena: Electrochem. Commun. 136 (2022) 107207. doi:10.1016/j.elecom.2022.107207
• 45)   D.  Wu,  K.  Kusada,  T.  Yamamoto,  T.  Toriyama,  S.  Matsumura,  I.  Gueye,  O.  Seo,  J.  Kim,  S.  Hiroi,  O.  Sakata,  S.  Kawaguchi,  Y.  Kubota and  H.  Kitagawa: Chem. Sci. 11 (2020) 12731–12736. doi:10.1039/D0SC02351E
• 46)   T.X.  Nguyen,  Y.-H.  Su,  C.-C.  Lin,  J.  Ruan and  J.-M.  Ting: Adv. Sci. 8 (2021) 2002446. doi:10.1002/advs.202002446
• 47)   S.  Nellaiappan,  N.K.  Katiyar,  R.  Kumar,  A.  Parui,  K.D.  Malviya,  K.G.  Pradeep,  A.K.  Singh,  S.  Sharma,  C.S.  Tiwary and  K.  Biswas: ACS Catal. 10 (2020) 3658–3663. doi:10.1021/acscatal.9b04302
• 48)   J.K.  Pedersen,  T.A.A.  Batchelor,  A.  Bagger and  J.  Rossmeisl: ACS Catal. 10 (2020) 2169–2176. doi:10.1021/acscatal.9b04343
• 49)   Z.  Jin,  X.  Zhou,  Y.  Hu,  X.  Tang,  K.  Hu,  K.M.  Reddy,  X.  Lin and  H.-J.  Qiu: Chem. Sci. 13 (2022) 12056–12064. doi:10.1039/D2SC04461G
• 50)   Z.  Jin,  J.  Lyu,  Y.-L.  Zhao,  H.  Li,  X.  Lin,  G.  Xie,  X.  Liu,  J.-J.  Kai and  H.-J.  Qiu: ACS Mater. Lett. 2 (2020) 1698–1706. doi:10.1021/acsmaterialslett.0c00434
• 51)   J.W.  Sturman,  E.A.  Baranova and  Y.  Abu-Lebdeh: Front. Energy Res. 10 (2022) 862551. doi:10.3389/fenrg.2022.862551
• 52)   A.  Sarkar,  L.  Velasco,  D.  Wang,  Q.  Wang,  G.  Talasila,  L.  de Biasi,  C.  Kübel,  T.  Brezesinski,  S.S.  Bhattacharya,  H.  Hahn and  B.  Breitung: Nat. Commun. 9 (2018) 3400. doi:10.1038/s41467-018-05774-5
• 53)   Z.  Lun et al.: Nat. Mater. 20 (2021) 214–221. doi:10.1038/s41563-020-00816-0
• 54)   R.  Zhang et al.: Nature 610 (2022) 67–73. doi:10.1038/s41586-022-05115-z
• 55)   T.  Löffler,  A.  Ludwig,  J.  Rossmeisl and  W.  Schuhmann: Angew. Chem. Int. Ed. 60 (2021) 26894–26903. doi:10.1002/anie.202109212
• 56)   H.  Chun,  E.  Lee,  K.  Nam,  J.-H.  Jang,  W.  Kyoung,  S.H.  Noh and  B.  Han: Chem Catal. 1 (2021) 855–869. doi:10.1016/j.checat.2021.06.001
• 57)   Y.  Yao et al.: Proc. Natl. Acad. Sci. USA 117 (2020) 6316–6322. doi:10.1073/pnas.1903721117
• 58)   T.A.A.  Batchelor,  T.  Löffler,  B.  Xiao,  O.A.  Krysiak,  V.  Strotkötter,  J.K.  Pedersen,  C.M.  Clausen,  A.  Savan,  Y.  Li,  W.  Schuhmann,  J.  Rossmeisl and  A.  Ludwig: Angew. Chem. Int. Ed. 60 (2021) 6932–6937. doi:10.1002/anie.202014374
• 59)   L.  Banko,  O.A.  Krysiak,  J.K.  Pedersen,  B.  Xiao,  A.  Savan,  T.  Löffler,  S.  Baha,  J.  Rossmeisl,  W.  Schuhmann and  A.  Ludwig: Adv. Energy Mater. 12 (2022) 2103312. doi:10.1002/aenm.202103312
• 60)   B.  Zhu,  S.  Alavi,  C.  Cheng,  H.  Sun,  H.  Zhao,  K.S.  Kim,  J.  Mostaghimi and  Y.  Zou: Adv. Energy Mater. 23 (2021) 2001116. doi:10.1002/adem.202001116
• 61)   L.  Chen,  Z.  Chen,  X.  Yao,  B.  Su,  W.  Chen,  X.  Pang,  K.-S.  Kim,  C.V.  Singh and  Y.  Zou: J. Mater. Inform. 2 (2022) 19. doi:10.20517/jmi.2022.23
• 62)   Z.  Lu,  Z.W.  Chen and  C.V.  Singh: Matter 3 (2020) 1318–1333. doi:10.1016/j.matt.2020.07.029
• 63)   A.A.H.  Tajuddin et al.: Adv. Mater. 35 (2023) 2207466. doi:10.1002/adma.202207466