Recent Progress in High Entropy Alloys for Electrocatalysts

Abstract

High entropy alloys (HEAs), which can incorporate five or more constituents into a single phase stably, have received considerable attention in recent years. The composition/structure complexity and adjustability endow them with a huge design space to adjust electronic structure, geometric configuration as well as catalytic activity through constructing reaction active sites with optimal binding energies of different reaction intermediates. This paper reviews the recent progress on the preparation methods, characterization techniques, electrocatalytic applications and functional mechanisms of HEAs-based electrocatalysts for hydrogen evolution, oxygen evolution and oxygen reduction reactions. The synthesis approaches for HEAs from bottom-up (high-energy ball milling, cryo-milling, melt-spinning and dealloying) to top-down strategies (carbothermal shock, sputtering deposition and solvothermal) and the corresponding materials characterizations are discussed and analyzed. By summarizing and analyzing the electrocatalytic performance of HEAs for diverse electrocatalytic reactions in water electrolysis cells, metal-air batteries and fuel cells, the basic principle of their designs and the relevant mechanisms are discussed. The technical challenges and prospects of HEAs-based electrocatalysts are also summarized with the proposed further research directions. This review can provide a beneficial theoretical reserve and experimental guidance for developing high performance electrocatalytic materials via the paradigm of high entropy.

Graphical Abstract

1 Introduction

In the exploration of clean and sustainable energy sources such as solar, wind, waterfall, etc. for electricity energy generation, the corresponding energy storage and conversion technologies have to be developed to smooth and store the generated weather-dependent intermittent electricity energy for practical applications. For this purpose, some advanced electrochemical energy technologies including water electrolysis to produce hydrogen, fuel cells, and metal-air batteries have been extensively researched and developed in recent years [1,2,3,4,5]. However, there are still some rooms for performance improvement of these electrochemical energy technologies in terms of energy/power densities and life-time. Specifically, the rates of electrocatalytic reactions at the electrodes are still insufficient for many practical applications. The electrocatalytic reactions are (1) oxygen evolution reaction (OER) at the anode and hydrogen evolution reaction (HER) at the cathode for water electrolysis cells; (2) oxygen reduction reaction (ORR) at the cathode and hydrogen oxidation reaction (HOR) at the anode for fuel cells; and (3) both ORR and OER at the anode for metal-air batteries. In these reactions, electrocatalysts are essential for speeding up the sluggish reaction kinetics and complex multi-electron transfer processes. Noble metal-based electrocatalysts, especially platinum (Pt)-based materials, have been recognized as the most viable catalytic materials. However, the scarcity and high cost of noble metals may limit their use in large-scale [6,7,8]. Therefore, low-cost and high-performance low-loading Pt and even non-Pt electrocatalysts have been explored for meeting the needs of rapid development. One of the current hot researches in this field is focused on alloying noble metals with earth-abundant 3D transition metals to decrease the amount usage of noble metal with an attempt to maintain a comparable or even better level of electrocatalytic performance. Among them, Pt-based alloys, mostly close to a stoichiometric ratio of Pt₃M (M is non-Pt metals), could exhibit around twofold enhancement in the Pt-mass activity [9,10,11,12,13,14,15,16,17,18,19,20]. These pioneer works provide important insights into the affecting mechanisms of alloying for activity enhancement: (i) ligand effects Due to the proximity of transition metals with different electron negativities from Pt, the direct electron interaction can happen over one to three atomic layers; and (ii) geometric effects The transition metals could shorten nearest-neighboring Pt–Pt interatomic distances in Pt alloys. Both of these mechanisms can induce a change of the electronic structure of Pt surface and therefore weaken the adsorption of intermediate oxygenated species.

Although some progress in controllable preparation and performance optimization of Pt-based alloys has been achieved, there are still shortcomings and challenges in this field:

• (i) The large particle size and low Pt utilization of the reported Pt-based alloys are not conductive to their practical applications. Therefore, morphology regulation with high specific surface area, such as nanoplates, nanowires, nanotubes, etc., is particularly important.

• (ii) The transition metals in the Pt-based alloys are easy to be oxidized and dissolved in the harsh electrochemical environment, leading to both losses of catalytic activity and stability, although most of the reported Pt-based alloys possess high electrocatalytic activity.

• (iii) Not all metal elements can form stable alloys, and there exists a co-solution boundary among some of them.

One type of the advanced metal alloys, high entropy alloys (HEAs), generally defined as solid solutions consisting of at least five principal metal components in an approximately equal atomic ratio [21], has come at the beginning of the twenty-first century [22,23,24] (Scheme 1). Due to their outstanding mechanical, catalytic, and anti-irradiation properties, they have been widely explored as the electrocatalysts for many electrochemical applications. Regarding the multicomponent alloys, Greer [22] pointed out that the more elements in the alloy, the easier to form amorphous alloy. Cantor et al. [23] prepared multicomponent alloys containing several metals via casting and melt spinning techniques. The obtained multicomponent transition metal alloy (FeCrMnNiCo) with an atomic radius difference less than 15% exhibited a surprising degree of inter-solubility in a single face-centered cubic (FCC) phase instead of forming an amorphous glassy structure, indicating the formation of a solid solution according to Hume-Rothery rule. At the same period, a series of research works on multicomponent alloys with equal and nearly equal atomic ratios were synthesized by Yeh’s group [24]. It was proposed that with the increment of the number of components in the alloy, their high mixing entropy could hinder the generation of intermetallic compounds and promote the formation of simple solid solution phases, based on which a concept of HEAs was putting forward. Because the number of constituents is five or more, the contribution of the mixed configuration entropy can be obtained through Eq. 1.

$$\Delta S_{{{\text{mix}}}} = - R\mathop \sum \limits_{i = 1}^{n} C_{i} \ln C_{i}$$

(1)

where, R is the molar gas constant (8.314 J K⁻¹ mol⁻¹), and C_i represents the molar fraction of the element component. So, ∆S_mix of the HEAs with an equal molar ratio for metallic elements in the liquid state or the solid-solution can be simplified into Eq. 2 [25, 26]:

$${\Delta }S_{{{\text{mix}}}} = - R\ln n$$

(2)

where, n represents the number of elements in the alloy. In general, HEAs can be characterized as an alloy that contains at least five elements with ∆S_mix greater than 1.61R. With five or more metal components incorporated into a single solid-solution phase, the large miscibility limitation of some conventional alloys will break down and allow for the continuous optimization of surface properties to maximize the reactivity of HEAs. Therefore, the research interest in HEAs over the past 20 years has been mainly focused on high-hardness and high-temperature alloy materials, anticorrosive coating, and photo-thermal conversion materials due to their remarkable properties such as outstanding mechanical strength, excellent corrosion resistance, as well as thermal and oxidation resistances under severe conditions [27,28,29,30,31].

Scheme 1

Major developments of HEAs

In recent years, HEAs as electrocatalytic materials have attracted increasing attention [32,33,34]. With the unique surface atomic configurations and electronic structure induced by the combination of multiple metal elements, which are different from that of pure metals and conventional binary and ternary alloys, HEAs can provide a more possibility for continuously adjusting the adsorption/desorption energy of reactants and intermediates on the catalyst surface. Briefly, HEAs have the following four effects that make them very suitable for application in the field of electrocatalysis:

• (i) High entropy effects. A high mixing entropy of HEA can reduce the Gibbs free energy of solid solution formation. It can more likely form single solid solution phase, instead of multiphase alloy structure, directly contributing to its long-term durability of HEAs.

• (ii) Lattice distortion effects. Because metallic atoms with different sizes randomly occupies the lattice position, severe lattice distortion and the consequent stain in an HEA can change the d-band center, which has a significant effect on the adsorption energy of reactants and intermediates [35,36,37,38,39].

• (iii) Sluggish diffusion effects. In an HEA, there is no smooth atomic diffusion channel due to the different sizes of various metal elements, which makes the diffusion coefficient of atoms in the surface low. This also contributes to the reliable stability of HEAs as electrocatalysts.

• (iv) Cocktail effects. By choosing appropriate alloy elements and optimizing the interaction between them, the HEAs with more than the average value of element properties can be obtained. This can provide the possibility for regulating and obtaining new HEAs with unexpected properties.

Thus, in this paper, recent progress in HEAs-based materials is comprehensively reviewed with focusing on their diverse designs, synthesis, functional mechanisms and applications in different electrochemical energy devices (Fig. 1). Moreover, to facilitate the further research and development, the technical challenges are analyzed and the possible research directions are also proposed for overcoming the challenges toward practical applications.

Fig. 1

Summary of the high entropy alloys (HEAs) from preparation, physicochemical characterizations to their electrocatalytic applications

2 Preparation Methods for HEA Materials

Regarding the preparation of HEA materials, diverse unique synthetic methods have been developed in the past years. They can be classified into two categories of top-down and bottom up approaches [40]. The former is simply used to fabricate HEAs materials by applying appropriate energy input to make various bulk pure metal materials into a single composite alloy phase. While the latter is generally employed to prepare HEAs nanoparticles based on the interaction of atoms or some molecular species through a set of chemical/physical reactions (Fig. 2). Herein, the recent development of preparation methods for HEAs materials are summarized. The general features of different methods and the basic characteristics of the as-prepared HEAs products such as constituent elements, crystallinity, micro-structural morphology will be discussed in the following subsections.

Fig. 2

Synthesis approaches of HEAs

2.1 Top-down Approaches

The top-down approach is generally a mechanical way to prepare nano-alloy materials. The general principle of this method is use the input energy to crush, mill and shape the bulk metals and make them reach the desired shapes and nanoscales. The main advantages of the top-down approach are the relative low cost, high productivity, easy operation and the ability to produce nanoscale HEAs [41,42,43]. Moreover, the further chemical/electrochemical dealloying processes can selectively etch part of transition metals or in-situ generate new active sites on the surface [44, 45]. These processes not only adjust the local electronic structure but also form porous structure to promote their electrocatalytic performance. However, the final HEA products usually possess poor substrate applicability and disorder morphology structure with defective surfaces.

Mechanical alloying and melt-alloying are the commonly used top-down approaches to produce functional HEAs materials for electrocatalytic applications.

The mechanical alloying strategy, mainly including ball-milling and cryo-milling [46, 47], is an effective method for non-equilibrium synthesis of fine-grained alloy powders. Through ball milling the metal powders with the process controlling agents, HEAs powder with different constituents and uniform particle size can be obtained. Moreover, this technique can induce chemical reactions that do not normally occur at room temperature. Fu et al. [48] successfully prepared self-supported CoCrFeNiAl HEAs electrocatalyst for HER by combing mechanical alloying with spark plasma sintering (SPS) consolidation (Fig. 3a). Through repeated strong collision and friction in a planetary ball miller under an argon atmosphere as well as post heat treatment crystallization, the alloy effect of Al/Cr with Co/Fe/Ni at an atomic level endowed CoCrFeNiAl HEA with high stability in acidic media. The as-prepared CoCrFeNiAl HEA after HF treatment and in situ electrochemical activation for 4 000 cycles of cyclic voltammetry presented superior HER activity, being attributed to the metal hydroxides/oxides groups on the surface of obtained electrocatalysts.

Fig. 3

a Schematic illustration of the fabrication process of CoCrFeNiAl HEA sheets by mechanical alloying and spark plasma sintering consolidation. Reproduced with permission from Ref. [48]. Copyright 2020, Elsevier B. V.; b mechanochemical synthesis of Pt/Ru-(NiMgCuZnCo)O entropy-stabilized metal oxide solid solution. Reproduced with permission from Ref. [49]. Copyright 2019, American Chemical Society

High-energy ball milling (HEBM), as a new technique initially developed for producing novel metastable materials, which cannot be synthesized by using thermal equilibrium processes, is quite different from the traditional ball milling technique [50]. It can induce chemical reactions that do not normally occur at room temperature. So it is also called reactive ball milling and used to produce nanosized metal oxides. Through the mechanical energy and frictional heating, Dai et al. synthesized a single-phase high entropy metal oxide (NiMgCuZnCo)O without additional thermal treatment. Also, noble metal oxides (PtO₂ or RuO₂) could be dispersed onto the rock salt lattice [Pt/Ru-(NiMgCuZnCo)O] on the nanometer scale, or even the single atom scale, by solid-sate grinding assisted by high entropy stabilization (Fig. 3b). The as-obtained (NiMgCuZnCo)O with 5 wt% (wt% is in the weight fractions) of Pt/Ru showed a high stability for CO₂ reduction reaction [49]. This could be due to that entropy-stabilized metal oxide solid solution (NiMgCuZnCo)O acted as an excellent support to stabilize highly dispersed even single atomically dispersed Pt or Ru, at temperatures up to 500 °C.

The cryo-milling strategy is the process of cooling or chilling materials by dry ice, liquid carbon dioxide or liquid nitrogen, and followed by grounding them into small sized particles. It is a variation of mechanical milling and can easily overcome the challenges in the conventional grinding process such as heat generation, introduction of tensile stresses, and less tool life clogging [51]. By using an easily scalable and precise composition control route of melting and cryo-milling technology to treat the cast HEAs precursor, Nellaiappan et al. [52] successfully prepared AuAgPtPdCu particles with a single FCC-faced crystalline structure and an average size of (16 $\pm$ 10) nm. Thanks to the uniform distribution of five metal elements, Cu atoms could be stabilized by other metals in FCC-facet crystalline structure as a “single-atom catalyst”, which could be efficiently used for the conversion of CO₂ with Faradic efficiency of about 100% toward gaseous products at a low applied potential (− 0.3 V vs. RHE).

The melt-alloying strategy is the most initial and commonest method to prepare HEAs, including vacuum induction melting [53,54,55], vacuum arc melting [56,57,58], and vacuum plasma beam melting [59,60,61]. Unfortunately, HEAs synthesized via melt-alloying usually suffer from small specific surface area, poor pore structure and inert surface, being unfavorable for electrocatalytic reactions. Dealloying strategy could selectively remove one or more components from the alloy via chemical or electrochemical method, and thus efficiently regulate the surface properties of the as-prepared bulk HEAs, such as pore structure, metal valence, coordination numbers of specific metal, and strain types, leading to enhanced electrocatalytic performance. Therefore, combing melting-alloying with dealloying process to prepare functional HEAs for electrocatalytic field is practical and attracting more and more attention. By taking advantages of that Al can be easily dissolved in acidic or basic media to achieve the goal of dealloying, researchers usually adopt Al as the main metal component to prepare Al-rich HEAs in an induction-melting furnace. Qiu et al. successfully prepared a series of nanostructure HEAs via melt-spinning dealloying strategy for electrocatalytic applications, including nanoporous AlNiCuPtPdAu [62] (Fig. 4b), ligament-pore structure AlNiCoIrMo [63] (Fig. 4c), and rugged AlNiCoRuMo nanowires with in-situ formed surface spinel oxide [64]. Similarly, instead of using NaOH solution to remove extra Al metal, Qin et al. [65] developed nanoporous noble metal Cu₃₀Au₂₃Pt₂₂Pd₂₅ quasi-HEA microspheres using HNO₃ solution to etch extra Cu metal, presenting a small ligament/hierarchical porous structure inside the grains with a high specific surface area of 69.5 m² g⁻¹ (Fig. 4a). Wang et al. [66] also fabricated porous CrMnFeCoNi HEAs by using H₂SO₄ solution to directly etch the commercial cantor alloy power, which exhibited remarkable catalytic activity with converting > 90% of p-nitrophenol into p-aminophenol. Yang et al. [67] employed three post-processing methods (annealing + dealloying, dealloying and melt spinning + dealloying) to treat the eutectic FeCoNiCrNb HEAs to obtain low cost bulk porous nanostructure with corrosion resistant high entropy intermetallic ligaments, which showed excellent OER activity and durability.

Fig. 4

a Synthesis process of nanoporous CuAuPtPd quasi-HEA microspheres combing arc-melting and dealloying strategies. Reproduced with permission from Ref. [65]. Copyright 2019, MDPI; b STEM-EDS mapping of senary nanoporous AlNiCuPtPdAu HEAs prepared by combing scalable alloying melting, fast cooling, and dealloying. Reproduced with permission from Ref. [62]. Copyright 2019, The Royal Society of Chemistry; c STEM-EDS mapping of quinary nanoporous AlNiCoIrMo HEAs by dealloying predesigned Al-based precursor alloys. Reproduced with permission from Ref. [63]. Copyright 2019, Wiley–VCH

It should be emphasized that the nanoscale dealloying is a conventional material surface modification technique to improve pore structure and enhance the electrocatalytic performance of alloyed materials, and it has been widely demonstrated in a typical M-rich Pt-M or Au-M bimetallic alloy. In other words, selectively (electro)chemical leaching of a less noble metal M from an M-rich Pt-M/Au-M alloy can give a “spongy” structure with ultrafine nanopores and nanoligaments or lead to a core–shell structure with a thin Pt/Au shell [68, 69], which could enhance the electrocatalytic activity and stability of alloyed particles [70,71,72,73]. The properties of large open surface areas accessible by reactant molecules, abundant defects such as steps and kinks, and catalytically active under-coordinated atoms on surfaces resulting from spongy structure are highly desirable for high-performance electrocatalysts. Surface lattice strain arising from the difference of core–shell composition can be successfully introduced into a thin Pt/Au metal shell, and the induced geometric surface lattice strain has been proposed as the means to tune the binding energy of reactive intermediates, leading to optimal overall catalyst performance (Fig. 5a) [74]. Hence, utilizing the dealloying strategy to fabricate HEAs with nanoporous spongy structure or ordered core–shell structure is a promising method for preparing ideal electrocatalysts and well-worth exploring in depth (Fig. 5b).

Fig. 5

a Compressive and tensile strained Pt surfaces are predicted to exhibit reduced and increased bond strengths with surface adsorbates like oxygen. Reproduced with permission from Ref. [74]. Copyright 2016, Elsevier B. V.; b schematic diagram of porous or core–shell structures formed after dealloying process for high entropy Pt-based alloy nanoparticles; c illustration of the evolution in size-dependent morphology and composition of dealloyed Pt-Cu and Pt-Co particle electrocatalysts. Reproduced with permission from Ref. [75]. Copyright 2012, American Chemical Society

In order to enhance the electrocatalytic performance of HEAs by dealloying, it is necessary to elaborate the dealloying mechanism of alloyed materials. Due to the difference of standard electrode potentials between less-noble and noble metals, the specific less-noble metal of alloyed materials can be selectively dissolved to form pits on the surface. Simultaneously, noble metal with local low coordination numbers starts to diffuse and agglomerate into cluster/island to lower down the surface energy [76, 77]. When the dissolution rate of less-noble metal is faster, more noble meal atoms will diffuse to the clusters and further get undercut into a spongy porous structure until noble metal atoms cover the whole surface of alloy to bring about passivation. In contrast, when the diffusion rate of noble metal dominates the dealloying process compared to that of less-noble metal, the rapid surface diffusion of noble metal will cover the surface of the particle quickly and cause fast annihilation of pit vacancies as well as decrease of surface energy, suppressing the formation of voids and pores, and in this way leading to a core–shell structure. Consequently, by regulating the rates of (electro) chemical dissolution for less-noble metal and surface diffusion for noble metal, the various structures of dealloyed materials can be obtained. Strasser et al. found that the particle size of alloy had a great effect on the basic dealloyed morphological and compositional structures in dealloyed nanoparticles [74, 75]. With the increment of particle size, the morphology of dealloyed nanoparticles changes from single non-porous core–shell nanoparticles to porous multiple core–shell particles (Fig. 5c) [75], demonstrating that the interfacial energy and the size-dependent diffusion rates of Pt are playing a key controlling role.

2.2 Bottom-up Approaches

Different from top-down approaches, the bottom-up synthetic approach is to prepare materials from atomic or molecular species via chemical reactions [78, 79], and the precious regulating of the nucleation and growth of precursors can lead to the production of good quality products with nano-structures and few specific metallurgical defects [80, 81]. On the other hand, the bottom-up approaches usually contain complex operation process, diverse solvents and special devices, accounting for the relatively high cost and low productivity.

Carbothermal shock strategies, sputtering deposition strategies and solvothermal strategies are the main reported bottom-up approaches to synthesize HEAs materials.

The carbothermal shock strategy is a facile high throughput method to prepare HEAs materials on carbon support by a rapid electrically joule heating and fast cooling (10⁵ K s⁻¹ to 2 000 K) of the sample impregnated with diverse metals precursors. By optimizing the operation parameters such as the metal precursor concentration, the shock duration, the heating/cooling rate, and the substrate, HEAs materials with variable composition, narrow particle size distributions and uniform dispersion can be obtained [82, 83]. However, for carbothermal shock strategy, it needs harsh conditions that the substrates call for good electrical conductivity and excellent thermal stability to avoid to be affected by extremely high temperature instantaneously as it utilizes the current pulse as the energy input for heating. Furthermore, because thermal shock could cause transient high supersaturation of the metal ions, it is difficult to control the thermodynamic process of various metals nucleation and growth, which is unfavorable for the precise regulation of particle size and constituents of as-prepared HEAs.

Hu et al [82] firstly adopted carbothermal shock method for synthesizing a wide range of HEAs nanoparticles with desired composition, sizes and phases on carbon nanofiber support (Fig. 6a–c). The ultra-high heating/cooling rate and surface defects with surface-bound oxygen (O*) on carbon nanofiber support can induce a rapid particle “fission” and “fusion” event, resulting in mixtures of multiple elements with uniform particle dispersion. The as-synthesized PtPdRhRuCe nanoparticles exhibited ~ 100% conversion of ammonia (NH₃) and > 99% selectivity toward NO_x (NO + NO₂) at a relatively low operation temperature of 700 °C. Li et al. [84] utilized the aligned carbon nanofibers (ACNFs) as the substrate for carbonthermal shock with a suitable CTS current direction. With this process, the quinary FeNiCoMnMg HEA-NPs/ACNFs electrodes were successfully prepared (Fig. 6d), which presented a high capacitance of 203 F g⁻¹ and a specific energy density of 21.7 Wh kg⁻¹. It was reported that HEA nanoparticles were uniformly in-situ fabricated onto carbon fibers while the current direction of loaded electric pulse was along the fibers rather than perpendicularity, manifesting that HEA-NPs formation on ACNFs had a significant relationship with the direction of the loaded current.

Fig. 6

a The synthesis of samples and the temporal evolution of temperature during the 55-ms thermal shock; b low-magnification and single-particle elemental mappings, an HAADF image, and the corresponding atomic mappings for a binary PtNi alloy; c elemental mappings of an HEA composed of eight dissimilar elements (Pt, Pd, Ni, Co, Fe, Au, Cu and Sn). The scale bar, 10 nm. Reproduced with permission from Ref. [82]. Copyright 2018, American Association for the Advancement of Science; d the formation schematic illustration of HEAs nanoparticles on carbon nanofibers substrates via carbothermal shock strategy. Reproduced with permission from Ref. [84]. Copyright 2020, Elsevier B. V.

The sputtering Deposition strategy is one of the most important physical vapor deposition (PVD) techniques, which involves ejecting material from metal targets onto a substrate. More specifically, argon is introduced into the chamber as the sputtering gas followed by its ionization. Under the action of electric field force, the relatively abundant Ar⁺ ions bombard onto the target surface. Meanwhile, the surface atoms of the target material are physically excited, detached from the target, and assembled on the substrate to form various required materials [85]. Sputtering deposition permits accurate regulation of the composition and uniformity of HEAs and great flexibility in the types of as-prepared materials via adjusting target composition, deposition rate, substrate temperature and so on. Yang et al. [86] developed magnetron sputtering deposition strategy to synthesize the quaternary FeCoCrNi film with carbon cloth as the substrate (Fig. 7a). XRD patterns and high-resolution TEM (HRTEM) images showed the FCC phase with interconnected nanocrystals with the size of ~ 5 nm. Benefited from the irreversible surface reconstruction during OER process, the highly oxidized Ni⁴⁺ species as active sites could be in-situ formed and favorable for electrocatalyzing OER. Schuhmann et al. [87, 88] developed a noble metal-free CrMnFeCoNi multinary element alloy, with a diameter of about (1.7 ± 0.2) nm, via combinatorial co-sputtering from elemental targets into an ionic liquid (IL) (Fig. 7b). The high-entropy effect promoted single solid solution formation. The IL with negligible vapor pressure served as a stabilizer and suspension medium due to its negligible vapor pressure. Followed by potential-assisted immobilization of the HEAs nanoparticles at an etched carbon nanoelecrode, the as-prepared quinary alloy system presented surprisingly high intrinsic activity toward ORR. It is worthwhile emphasizing that the potential of multinary alloys toward tailoring catalytic properties by a high-entropy-induced formation of a single solid solution phase with homogeneous distribution of all constituents can provide a big number of novel active sites resulting from this mixture.

Fig. 7

a Schematic illustration of the pristine FeCoCrNi film preparation via the sputtering deposition method, corresponding structural characterizations, the SEM image, the magnified image, the HRTEM image, and the 3D atomic-probe tomography probing reconstruction image. Reproduced with permission from Ref. [86]. Copyright 2020, Springer Nature; b strategy for the evaluation of HEA NPs by means of combinatorial co-sputtering into an ionic liquid. Reproduced with permission from Ref. [87]. Copyright 2018, Wiley–VCH

The solvothermal strategy, as one of the commonest and the most effective synthetic routes to fabricate functional HEAs with controlled sizes and morphology [89], has been developed on the basis of the hydrothermal method. In this strategy, the original mixture is reacted in a closed system such as autoclave with organic or non-aqueous solvent as solvent under certain temperature and autogenous pressure of solutions. It differs from hydrothermal reaction in that the used solvent is organics instead of water. By regulating types of metal precursors, reductants, surfactants, reaction temperature and time, HEAs with controlled constituent composition, morphology structure, crystalline phases and narrow particle size distribution can be obtained (Fig. 8a). Although the solvothermal strategy suffers from complex operation steps, toxic reagents and high cost, it is still a hotspot in the research field of HEAs preparation. Wang et al. [90] successfully prepared porous tetrametallic PtCuBiMn nanosheets by a one-step solvothermal method in a formamide solvent system, combining the concepts of crystal symmetry, shape control and acid etching. The synthetic procedure for PtCuBiMn nanosheets involved the co-reduction of Pt(acac)₂, Cu(acac)₂, Mn(acac)₂ and BiNO₃ in the presence of formamide, polyvinylpyrrolidone (PVP), and KI at 130 °C for 3 h. The as-synthesized PtCuBiMn nanocrystals possessed porous and nanosheet-like structures with an average size of ~ 15 nm and thickness of ~ 4 nm. Furthermore, PtAgBiCo triangular nanoplates with a high morphological yield (> 90%) was synthesized via a two-step solvothermal route with a gel-like materials as template [91]. Recently, Wang et al. [92] fabricated PtNiFeCoCu HEAs nanoparticles through a simple low-temperature oil phase strategy at atmospheric pressure (Fig. 8b). With (1-Hexadecyl) trimethylammonium chloride as the surfactant, glucose and oleylamine as the reductant, Mo(CO)₆ as the structure-directing reagent, M(acac)₂ (M represents Pt, Ni, Fe, Co, Cu, respectively) precursors could be co-reduced to form uniform and ultrasmall particles with a diameter of (3.4 $\pm$ 0.6) nm, exhibiting excellent bi-functional electrocatalytic properties for both HER and methanol oxidation reaction (MOR). According to the authors, the use of acetylacetonates is the key point to the fabrication of HEAs, because the strong metal-acetylacetonate interaction could facilitate co-precipitation by slowing down the precipitation rate. Instead of using one pot method to synthesize HEAs materials, Kitagawa et al. [93] synthesized of novel solid-solution alloy nanoparticles in the immiscible alloy system such as Ag-Rh (Fig. 8c). Due to the fact that the redox potentials of the metal ions (Ag and Rh) are significantly different, the reduced Rh and Ag atoms do not mix spontaneously with each other and always form segregated structures. To address this challenge, the reductant solution is firstly heated and then the metal precursors at a low concentration are slowly added. Both Ag and Rh ions are simultaneously and rapidly reduced and the difference in reduction speed seems to be negligible. Based on this strategy, Kitagawa [94, 95] et al. also successfully synthesized RuRhPdIrPt and RuRhPdOsIrPt platinum-group-metal based HEAs nanoparticles in single phase solid solution, which presented remarkable electrocatalytic activity for HER and MOR, respectively.

Fig. 8

a Schematic diagram of HEAs nanoparticles synthesis via the solvothermal co-reduction method; b TEM, HRTEM images and corresponding elemental mappings of PtNiFeCoCu via the solvothermal co-reduction method with acetylacetonate as metal precursors. Reproduced with permission from Ref. [92]. Copyright 2020, Springer Nature; c schematic illustration of the synthesis of AgRh solid-solution alloy nanoparticles. Reproduced with permission from Ref. [93]. Copyright 2020, Wiley–VCH

3 Predictions, Characterizations and Strain Calculations of HEAs

3.1 Thermodynamic Parameter Calculations for HEAs Prediction

In order to describe the comprehensive effect of the atomic size difference in n-element contained alloy, the parameter $\delta$ can be expressed as follows (Eqs. 3 and 4) [96]:

$$\overline{r} = \mathop \sum \limits_{i = 1}^{n} C_{i} r_{i}$$

(3)

$$\delta = \sqrt {\mathop \sum \limits_{i} C_{i} \left( {1 - \frac{{r_{i} }}{{\overline{r}}}} \right)^{2} }$$

(4)

where, C_i is the atomic percentage of the ith component, r_i is the atomic radius, and $\delta$ is the mean square deviation of the atomic size of elements. It can be found that the large atomic size difference will cause serious lattice distortion and result in the increase in free energy in alloy, which leads to poor stability of solid-solution.

The enthalpy of mixing for the multi-component alloy system with n elements can be calculated from Eq. 5 [97]:

$$\Delta H_{{{\text{mix}}}} = \mathop \sum \limits_{i = 1 i \ne j}^{n} \Omega_{ij} C_{i} C_{j}$$

(5)

where, $\Omega_{ij}$ is the regular solution interaction parameter between the ith and the jth elements; C_i and C_j are the atomic percentages of the ith and the jth components, respectively. It is known that the negative $\Delta H_{{{\text{mix}}}}$ is inclined to make the different elements combine to form intermetallic compound, while the positive $\Delta H_{{{\text{mix}}}}$ means the less miscibility of different elements in the liquid alloy, which leads to separation or segregation of different elements in alloy [98]. Hence, the larger absolute value of $\Delta H_{{{\text{mix}}}}$ will make solid-solution form harder.

For a multi-component system, it is difficult to calculate ${\Delta }G{ }$(the difference in Gibbs free energy between solid and liquid state) accurately at a certain composition and temperature. Takeuchi et al. [97] proposed an assumption for multi-component alloy systems: ${\Delta }G{ }$ at a certain composition is proportional to the free energy ${\Delta }G_{{{\text{mix}}}} { }$ of the liquid phase. This assumption can be successfully adopted to calculate the critical cooling rate of metallic glasses (Eq. 6)

$${\Delta }G_{{{\text{mix}}}} = {\Delta }H_{{{\text{mix}}}} - T{\Delta }S_{{{\text{mix}}}}$$

(6)

where, ${\Delta }H_{{{\text{mix}}}}$ is the enthalpy of mixing, ${\Delta }S_{{{\text{mix}}}}$ is the entropy of mixing, and T is the absolute temperature.

In addition to solid-solution phases, multi-component HEAs may also be formed as intermetallic compounds and amorphous phases [99,100,101]. If solid-solution phase possesses the lowest ${\Delta }G_{{{\text{mix}}}}$ among all possible formed phases, it will be the most possible to form solid solution phase during solidification. Simple solid-solution phases are expected to form in multi-component HEAs owing to their promising properties [24, 102]. Unfortunately, it is almost impossible to calculate ${\Delta G}_{{{\text{mix}}}}$ of all possible formed phases for a multi-component alloy. Hence, a simple approach to predict solid solution formation is necessary. The ratio between $T{\Delta }S_{{{\text{mix}}}}$ and ${\Delta }H_{{{\text{mix}}}}$ can be adopted to estimate the solid-solution formation ability (Eqs. 7 and 8) [98].

$$T_{{\text{m}}} = \mathop \sum \limits_{i = 1}^{n} c_{i} (T_{{\text{m}}} )_{i}$$

(7)

$$\Omega = \frac{{T_{{\text{m}}} \Delta S_{{{\text{mix}}}} }}{{\left| {\Delta H_{{{\text{mix}}}} } \right|}}$$

(8)

where, c_i is the atomic percentage of the ith component, (T_m)_i is the melting point of the ith component of alloy (K). $\Omega$ is defined as a parameter of ${\Delta }S_{{{\text{mix}}}}$ timing the average melting temperature of the elements over ${\Delta }H_{{{\text{mix}}}}$.

By comparing the value of $T_{{\text{m}}} \Delta S_{{{\text{mix}}}}$ and $\left| {\Delta H_{{{\text{mix}}}} } \right|$, the competitive relationship between $\Delta S_{{{\text{mix}}}}$ and ${\Delta }H_{{{\text{mix}}}}$ at a certain temperature can be clearly elaborated. The effect of $T_{{\text{m}}} \Delta S_{{{\text{mix}}}}$ can balance that of $\left| {\Delta H_{{{\text{mix}}}} } \right|$ for forming solid-solution phase. In another words, the driving force is equal to the resistance. Briefly, under the condition of $\delta \leqslant$ 6.6%, − 11.6 < ${\Delta }H_{{{\text{mix}}}}$  < 3.2 (kJ mol⁻¹) (Fig. 9a) and $\Omega \geqslant$  1.1, a solid solution HEA could be formed. Consequently, it is possible to predict the formation of solid solution HEAs according to their composition [103,104,105]. Taking the catalytic decomposition of NH₃ as a research target, among the already reported unary metal catalysts (Fig. 9b), Ru shows the best performance, while Ni is very active as a non-noble catalyst. Due to different metals having different reactivity, HEAs composed of multiple elements can continuously tune the surface structure and chemistry for optimized catalytic performance. With the assistance of the computational strategy, the large composition space among active metals (Ru, Rh, Co, Ni, Ir, Pd, Cr, Fe, Cu, and Mo) was explored. Firstly, a composition prescreening process to identify solid solution structures based on the above-mentioned phase formation conditions derived in well-studied bulk HEAs was developed. With these criteria, the composition space of 10 active elements (Ru, Rh, Co, Ni, Ir, Pd, Cr, Fe, Cu, and Mo) was explored. As shown in Fig. 9c, for the ternary alloys derived from 10 active elements (7 740 kinds of composition), the statistical data show that ~ 61% of the ternary composition is alloys (yellow dots), while the others are intermetallic, phase-separated, or amorphous structures (purple dots). With more elements in the multi-elemental system, the quantity of total screened composition increases exponentially to > 7 million (Fig. 9d). In addition, the ratio of the alloy phases in the screened composition increased steadily with the increment in elements (i.e., increasing mixing entropy), indicating a strong entropy-driven single-phase stabilization. This prescreening process covers the whole composition map of the 10 elements studied and helped us to identify potential composition that could lead to alloy phases well before synthesis.

Fig. 9

a $\delta$–${\Delta }H_{{{\text{mix}}}}$ plot delineating the phase selection in high entropy alloys. The dash-dotted regions highlight the individual region to form solid solutions, intermetallic compounds and the amorphous phase. Reproduced with permission from Ref. [98]. Copyright 2012, Elsevier B. V. b Reaction efficiency of unary metal catalysts for NH₃ decomposition as a function of nitrogen desorption energy, 1 kcal mol⁻¹ = 4.186 kJ mol⁻¹; c ternary alloy screening from the 10 active elements (Ru, Rh, Co, Ni, Ir, Pd, Cr, Fe, Cu, and Mo) at 1 500 K synthesis with composition of each element ranging from 5% to 50% at a 5% step size (i.e., 7 740 kinds of composition). Yellow, uniform solid solution phases; purple, intermetallic, phase-separated, or amorphous structures; d quantities of the composition (left) and the ratio of the alloy phase in the composition (right) as a function of the multi-elemental systems, showing a total of  >  7 million kinds of composition and the steady increase in the alloy ratio due to entropy stabilization. Reproduced with permission from Ref. [103]. Copyright 2020, American Association for the Advancement of Science

Besides, the stability of FCC and body centered cubic (BCC) solid solutions is well-delineated by the valence electron concentration (VEC, the number of total electrons including the d-electrons accommodated in the valence band), which provides valuable input to design crystal structures of HEAs [106]. Specifically, VEC is the physical parameter to illustrate the phase stability for FCC or BCC solid solutions. FCC phases are stable at a VEC value higher than 8, while BCC phases are stable at a value lower than 6.87. VEC can be calculated according to Eq. 9:

$${\text{VEC}} = \mathop \sum \limits_{i = 1}^{n} c_{i} ({\text{VEC}})_{i}$$

(9)

where, $c_{i}$ and $({\text{VEC}})_{i}$ are the atomic percentages and VEC of the ith component, respectively.

3.2 Characterizations of HEAs

In order to identify whether the as-obtained materials are HEAs or not, precise characterizations of the elemental types, contents and distributions are needed. Energy dispersive X-ray spectroscopy (EDS), as one of the widely used analytical technique for the characterization of elemental types and contents in as-prepared materials, has recently gained significant importance regarding its application to the chemical analysis of nanoparticles, especially in conjunction with the use of scanning electron microscope (SEM) and transmission electron microscope (TEM). With the help of highly sensitive EDS detectors, qualitative information on elemental types and contents with the spatial resolution of about 10 nm can be achieved within the same scanned area as provided by the electron microscope [107]. Inductively coupled plasma mass spectrometry (ICP-MS) and inductively coupled plasma optical emission spectroscopy (ICP-OES) are also powerful analytical techniques for elemental characterizations and generally allowing rapid sample throughput [108]. X-ray photoelectron spectroscopy (XPS) is another quantitative technique for elemental composition and chemical state analysis in near surface region of materials through the detection of the binding energy of photoelectrons. With the assistance of these techniques, elemental types and contents in as-prepared materials can be precisely quantified. A homogeneous element distribution without elemental segregation or phase separation is a convincing proof for the formation of HEAs. Combing STEM (Scanning Transmission Electron Microscope)-EDS elemental mapping with line scan profile, the complete and uniform mixing of multi-metal elements within the individual nanoparticles can be verified (Fig. 10a–b) [109]. On the whole, it is difficult to identify whether the as-synthesized materials are HEAs or not only with one single characterization method, so the combination of various techniques is necessary.

Fig. 10

Morphological and compositional characterization of PdCuPtNiCo HEAs/C: a STEM-EDS elemental mapping images and b line scan profile; c XRD patterns; d TEM image; e STEM-HADDF image. Reproduced with permission from Ref. [109]. Copyright 2021, The Royal Society of Chemistry

Once it is demonstrated that the as-prepared materials are HEAs combing various above-mentioned characterization techniques, it is necessary to identify the crystallographic structure of the as-synthesized HEAs. Because HEAs can possess different crystallographic structures, e.g., single-phase solid solution (FCC, BCC and Hexagonal Closepacked Structure (HCP)), or amorphous structures, which have a significant effect on their electrocatalytic performance. X-ray diffraction (XRD) technique has been widely applied for phase identification of crystalline materials, providing information including lattice constants, crystal facets and crystallite size on unit cell dimensions [110]. As shown in Fig. 10c, by comparing the diffraction patterns of pure metals with corresponding alloy, no peaks can be assigned to pure Pd, Cu, Pt, Ni, or Co, demonstrating the formation of a single-phase alloy. According to Bragg’s law and the relationship between d-spacing and the unit cell parameter a for different crystal structures, the obtained crystal structure can be assigned to FCC structure [92, 100]. Besides, the high-resolution TEM (HRTEM) image can further demonstrate the formation of single-phase and single-crystalline nanoparticles (Fig. 10d–e). The above characterizations both help to prove the successful synthesis of single-phase FCC HEAs.

SEM, TEM, and high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) are the common methods for imaging the micro morphology of materials. Recently, the advanced in-situ characterization technologies are used in the field of HEAs, aiming to observe how the metal precursors form HEAs under specific reaction conditions. Shahbazian-Yassar et al. [111] investigated the local structural evolution during the reduction reaction of oxidized FeCoNiCuPt HEA nanoparticles in atmospheric pressure H₂ environment (Fig. 11). As demonstrated, the oxide layer transformation into porous structures is observed as H₂ penetrates into the oxide layer but the reaction front mostly remains at the external surface of the oxide. Cu oxide can be fully reduced and further segregated into Cu nanoparticles, while Fe, Co, and Ni remain in the oxide phase. Further expansion of the oxide layer and size reduction of HEA core during H₂ reduction are observed, resulting in segregated HEA core and oxide shell. Obviously, the in-situ TEM characterization facilitates an in-depth understanding of HEA reduction mechanisms and provides a theoretical basis for designing high corrosion resistance and durable catalysts.

Fig. 11

Schematic showing the reduction of oxidized HEA nanoparticles at 400 °C. The oxidized HEA nanoparticles have a Pt rich HEA core with disordered oxide layers containing Fe, Co, Ni, Cu. During reduction in H₂, the HEA core shrinks due to transition metals’ outward diffusion, with expansion of the oxide and segregation of oxidized Cu into reduced Cu nanoparticles. Reproduced with permission from Ref. [111]. Copyright 2021, American Chemical Society

3.3 Strain Calculations

For HEAs, large mixing entropy not only promotes the formation of a stable single-phase solid solution structure, but also causes serious lattice distortion. Because of the large difference of individual elements in the atomic radius and the lattice constant, the strained surface on the HEAs can be formed. Such strained surface presents electrochemical properties that are significantly different from those of pure metal surface, also known as the “strain effect”. Due to the diversity and complexity of elements in HEAs, the strain quantifications of HEAs have not been reported yet, though many published papers about the strain calculation of binary alloys have been reported. According to the definition, the strain of binary alloy, $s\left( {{\text{alloy}}} \right)$, can be obtained through Eq. 10 [112]:

$$s\left( {{\text{alloy}}} \right) = \frac{{a_{{{\text{alloy}}}} - a_{{{\text{bulk}}}} }}{{a_{{{\text{bulk}}}} }} \times 100\%$$

(10)

where a_bulk is the lattice parameter of bulk metal (M_b) and a_alloy is the lattice parameter of the corresponding alloy materials (M_a). According to Bragg’s Law, the lattice parameters for M_a and M_b can be obtained by PXRD (Powder X-ray Diffraction, PXRD) data. For cubic systems, the lattice parameter $a$ can be obtained through Eqs. 11 and 12:

$$2d\sin \theta = n\lambda$$

(11)

$$d_{hkl} = \frac{a}{{\sqrt {h^{2} + k^{2} + l^{2} } }}$$

(12)

where, d is the crystal spacing, $\theta$ is the diffraction angle, n is an integer number, $\lambda$ is the wavelength of the X-ray, and (h, k, l) is the miller indices.

Similarly, by utilizing the EXAFS (Extended X-ray Absorption Fine Structure), the atomic distance between two metal atoms in the alloy materials, $R_{{{\text{alloy}}}}$, can be obtained, which can be used to estimate the lattice strain by comparing the atomic distance in pure bulk metal ($R_{{{\text{bulk}}}}$) (Eq. 13).

$$s\left( {{\text{alloy}}} \right) = \frac{{R_{{{\text{alloy}}}} - R_{{{\text{bulk}}}} }}{{R_{{{\text{bulk}}}} }} \times 100\%$$

(13)

In brief, when $s\left( {{\text{alloy}}} \right)$ is positive, the material is under tensile strain. On the contrary, it is under compressive strain when $s\left( {{\text{alloy}}} \right)$ is negative.

4 Electrocatalytic Applications of HEA Materials

Recently, the diverse potential applications of HEAs in heterogeneous electrocatalytic reactions, such as HER, OER and ORR have been greatly explored. HEAs can provide a large number of adjustable surface states (element configurations, strain) as novel active sites compared to conventional monometals, binary or even ternary alloy. Hence, HEAs could exhibit remarkable possibilities for optimizing the adsorption energy of reactant molecules and intermediates, finally endowing excellent electrocatalytic performance (Fig. 12). Besides, the high-entropy effect and sluggish diffusion effect can also enhance the long-term stability to fulfil actual requirements.

Fig. 12

Comparison between HEA electrocatalysts and traditional alloyed electrocatalysts

4.1 HER Application

Hydrogen energy development continues to be an important research and demonstration pathway for major economies around the world, which offers potentially significant advantages in terms of low or zero emissions and flexibility in fuel sources [113]. In hydrogen production through water electrolysis, HER is the necessary reaction at the cathode of the electrolysis cell. HER involves the reduction of protons to gaseous H₂. Its first step is the reduction of a proton on an active site (the Volmer step, Eq. 14) to form the adsorbed hydrogen atom (H*), followed by H₂ evolution, either through a second proton/electron transfer (the Heyrovsky step, Eq. 15) or the recombination of two H* (the Tafel step, Eq. 16) [114]. Among them, the Volmer step is generally considered to be the rate-limiting step for HER [115], for which, highly effective electrocatalysts are urgently needed.

$${\text{H}}^{ + } + {\text{e}}^{ - } + {* } = {\text{ H}}^{*} \quad \left( {\text{the Volmer step}} \right)$$

(14)

$${\text{H}}^{*} + {\text{H}}^{ + } + {\text{e}}^{ - } = {\text{ H}}_{2} \quad \left( {\text{the Heyrovsky step}} \right)$$

(15)

$$2{\text{H}}^{*} = {\text{ H}}_{2} \quad \left( {\text{the Tafel step}} \right)$$

(16)

Noble metal-based HEAs are widely used as electrocatalysts for HER because of their unique capabilities of fine-tuning adsorption energy, good long-term stability and abundant lattice distortion (Fig. 12). Taking advantage of the acoustic cavitation phenomenon in the ultrasonication process (Fig. 13a), Dai et al. [116] prepared PtAuPdRhRu HEAs nanoparticles with a diameter less than 3 nm supported on XC-72 carbon in a one-step process at room temperature. The as-prepared HEA-NPs/carbon displayed excellent HER activity with the onset potential of − 0.025 V (vs. RHE) and a Tafel slope of 62 mV dec⁻¹, which was smaller than that of the commercial Pt/C (77 mV dec⁻¹). This superior performance could be attributed to high-entropy at the nanoscale and strong synergistic effects between active metals. Besides, non-noble metal-based HEAs also exhibited outstanding HER activity comparable to the commercial Pt/C. By mechanical alloying and spark plasma sintering consolidation, Fu et al. [48] developed a self-supported CoCrFeNiAl for highly efficient HER in acidic condition with an overpotential of 73 mV to reach a current density of 10 mA cm⁻² after HF treatment and in situ electrochemical activation (Fig. 13b). According to the authors, the porous structure, exposed nanophases and abundant metal hydroxides/oxides on the surface of the electrocatalyst account for the superior performance.

Fig. 13

a Schematic illustration of HEA-NPs/carbon (PtAuPdRhRu supported on XC-72 carbon) synthesis and their application in HER. Reproduced with permission from Ref. [116]. Copyright 2019, Wiley–VCH; b polarization curves and Tafel slopes of as-prepared HEAs and 20 wt% Pt/C for HER. Reproduced with permission from Ref. [48]. Copyright 2020, Elsevier B.V

4.2 OER Application

OER, as the other half reaction of water electrolysis, plays an important role in the development of economical, highly efficient and eco-friendly energy storage and conversion systems. However, the sluggish kinetics of OER, involving multi-step electron transfer paths and large overpotential, makes it a limiting reaction for the whole catalytic process. Hence, developing efficient electrocatalysts for OER is the key to the advancement of a number of renewable energy technologies, involving water electrolyzers, solar fuels production and rechargeable metal-air batteries [117,118,119,120,121,122,123,124,125,126,127,128]. Different from HER, electrocatalytic OER needs to undergo four electron transfer process, and the reaction mechanism is more complex. As shown in Fig. 14a, oxygen-containing species (OH⁻ or H₂O) are firstly adsorbed onto catalytic active sites to form M–OH intermediates. Next, they react with OH⁻ or undergo the deprotonation process to form M–O species. And then, there are two reaction pathways: one is that two intermediates (M–O) react directly to form O₂; another one is that M–O intermediates react with oxygen-containing species (OH⁻ or H₂O) to form M-OOH species, further releasing O₂. The rate-determining step of OER can be judged by the Tafel slope. A smaller Tafel slope represents more outstanding OER catalytic kinetics. However, due to the complexity of the reaction process, the error of experimental estimation is usually large. So, Density Functional Theory (DFT) calculation is mainly used to predict rate-determining step of OER at present. As shown in Fig. 14b, the relationship between the electrocatalytic activity of as-prepared materials and free energy difference of intermediates (ΔG_O* − ΔG_HO*) exhibits the volcanic type. The material at the top of the volcanic type possesses the best OER performance.

Fig. 14

a OER mechanism for alkaline (the green route) and acidic (the red route) conditions. The black line indicates the formation of a peroxide (M–OH intermediate, while the light blue line indicates the direct reaction of two adjacent oxo (M–O) intermediates to produce oxygen. b The negative values of theoretical overpotential are plotted against the standard free energy of ΔG_O* − ΔG_HO*. Reproduced with permission from Ref. [129]. Copyright 2020, Wiley–VCH

Recently, it is widely reported that high-entropy oxides (HEOs) or hydroxides represent a promising catalyst for OER with earth-abundant and low-cost elements as main constituent composition. The proper incorporation of suitable metals into FeCoNi to form HEOs can modify the catalyst’s electronic states, leading to an optimum OOH*/O* adsorption on the oxide surface to enhance OER performance [42, 130]. Qiu et al. [45] developed a simple dealloying technology to prepare AlNiCoFeX (X = Mo, Nb, Cr) with the naturally formed HEOs surface, i.e., nano-porous HEA/HEO (np-HEA/HEO) (Fig. 15a). Combining a melt-spinning method to prepare the Al-based precursor alloys with further chemical etching in 0.5 M NaOH (1 M = 1 mol L⁻¹) solution, the as-prepared np-HEA/HEO achieved an OER overpotential of 240 mV to reach 10 mA cm⁻² and simultaneously with an enhanced electrochemical durability even after 2 000 cycles. The electron transfer from Fe to Ni will make the binding energy of Ni²⁺ negatively shift and the addition of Mo can enhance the electron transfer effect, thus modify the catalyst’s electronic states, leading to an optimum O* adsorption and facilitating the OER process. Similarly, by using melt-spinning and dealloying strategy, Yang et al. [67] successfully fabricated FeCoNiCrNb HEA/HEO with amorphous HEOs ultrathin films wrapped around the nanosized intermetallic ligaments. The intermetallic-oxide core–shell structure with diversity of the valence states of the oxides/hydroxides provides tremendous active sites and enlarged active surface area, leading to the ultrafast kinetics and long-term durability during OER.

Fig. 15

a FCC high-entropy (oxy)hydroxides with nanoporous structure for electrocatalyzing OER. Reproduced with permission from Ref. [45]. Copyright 2019, American Chemical Society; b electrochemical OER activity evaluations of various catalysts; c mass activities and TOFs of various catalysts at an overpotential of 300 mV; d ex situ SXAS measurements of various catalysts: Ni L-edge and O K-edge. Dash and solid lines represent the spectra collected at the open circuit condition and the operated potentials; f schematic illustration of the proposed overall OER pathway for the electrochemical activated FeCoCrNi film (EA-FCCN) catalyst. The black and blue oxygen atoms represent lattice and adsorbed oxygen, respectively. Reproduced with permission from Ref. [86]. Copyright 2020, Springer Nature

The electrochemical cyclic voltammetry activation is also used to prepare HEOs. Chai et al. [86] prepared the FeCoCrNi HEA film coated on carbon cloth by magnetron sputtering. After CV activation in 1.0 M KOH solution for 200 cycles (the potential region—0 to 0.6 V vs. Hg/HgO, the scanning rate—100 mV s⁻¹), the electrochemically activated FeCoCrNi film (EA-FCCN) catalyst delivered an excellent mass activity and turnover frequency (TOF) of 3601 A g_metal⁻¹ and 0.483 s⁻¹ at an overpotential of 300 mV in the alkaline electrolyte, respectively (Fig. 15b–c). A Fe–Ni dual-site with highly oxidized Ni⁴⁺ species is proposed to be the ultimate catalytic center for high OER activity (Fig. 15d). With these results, an overall illustration towards OER on (FeCoCrNi)OOH model to rationalize the enhanced OER activity of corresponding EA-FCCN catalyst is depicted in Fig. 15e. The interatomically electron interplay plays a significant role in optimization of the reaction pathways. Ni⁴⁺ species are favorably formed through a multistep evolution (Ni²⁺ → Ni³⁺ → Ni⁴⁺) induced by the electronic modulation through the Fe–O–Ni moiety and pre-adsorbed μ-O_aH species at the Ni-Co site, introducing holes into oxygen ligands to evoke the lattice oxygen activation mechanism (LOM) pathway, then driving the construction of Fe–Ni dual-site as ultimate catalytic center to loop the water oxidation. As a result, benefiting from the favorable formation of Ni⁴⁺ species and dynamically constructed Fe–Ni dual-site, EA-FCCN catalyst offers low overpotential and superb activity for OER.

4.3 ORR Application

ORR is a key reaction of energy conversion device, such as fuel cells and metal-air batteries, which is known to have sluggish reaction kinetics because of its multi electron transfer process. The following two overall mechanisms of ORR have been suggested: a direct four-electron pathway, in which O₂ is directly reduced to H₂O without the formation of hydrogen peroxide (H₂O₂) intermediate (Eq. 17) and a series two-electron pathway in which O₂ is reduced to H₂O via H₂O₂ (Eqs. 18–19) [131] (Fig. 16a). It is noteworthy that ORR proceeds through a 2e⁻ or 4e⁻ pathway depends on the dissociation barrier of the initial adsorbed O₂ on the electrode surface, which is also one of the essential foundations to design ideal electrocatalysts. Nørskov et al. [132] provided the ORR activity as a function of the calculated O* binding energy (∆Eo) via DFT calculations. As shown in Fig. 16b, for the pure metals on the left of the volcano plot, their electrocatalytic ORR activity improves with the increment of ∆Eo. This can be attributed to the ORR activity in this area with strong O₂ bonding, which is determined by the removal rate of O, OH and some oxygen-containing intermediates on the electrode surface. On the contrary, the pure metals on the right side of volcano plot exhibits a fading ORR activity with the increment of ∆Eo. This is due to the fact that the increment of O binding energy means the weaker adsorption of the O₂ molecule on electrode surface, leading to a reduced activity. Obviously, the binding strength between the catalysts and the adsorbed reactants, intermediates and products should be neither too strong nor too weak. The closer to the volcano top, the more moderate the binding energy of oxygen-containing species and the better the ORR activity.

Fig. 16

a Schematic diagram of the ORR mechanism; b ORR activity for a series of pure metals plotted against O* binding energy. Reproduced with permission from Ref. [133]. Copyright 2020, The Royal Society of Chemistry

Therefore, electrocatalysts with appropriate adsorption energy of O* or OH* to achieve both superior ORR activity and 4e⁻ selectivity are urgently needed for high output of new energy conversion devices. With the capability of forming single phase solid solutions from five or more constituents, HEAs can generate an immense number of possible active sites, which covers a wide range of adsorption energy to reach the top of volcano and further obtains optimal binding energy and reaction activity.

$${\text{O}}_{2} + 4{\text{H}}^{ + } + 4{\text{e}}^{ - } \to 2{\text{H}}_{2} {\text{O}}$$

(17)

$${\text{O}}_{2} + 2{\text{H}}^{ + } + 2{\text{e}}^{ - } \to {\text{H}}_{2} {\text{O}}_{2}$$

(18)

$${\text{H}}_{2} {\text{O}}_{2} + 2{\text{H}}^{ + } + 2{\text{e}}^{ - } \to 2{\text{H}}_{2} {\text{O}}$$

(19)

Schuhmann et al. [88] theoretically elucidated the correlation of adsorption energy distribution patterns (AEDP)-activity correlation and further identified it by the obtained experimental activity curves. Since there exist many active site centers (each active site represents an adsorption peak) in multi-component alloys and each will contribute one exponentially increasing current curve of different activity governed by the relative position of each adsorption peak maximum to the optimal binding energy, the authors simplified the number of active sites to 7 within one adsorption peak (Fig. 17a). The overall current response still exhibited one consistent exponential increase peak, followed by a plateau current once active site limitation is reached. The activity is governed by the position of the peak maximum regarding optimal binding energy (Fig. 17b). From the left-top of Fig. 17c, it is found that only one fixed and distinct binding energy exists for an elemental catalyst, corresponding a single exponential increase current curve. However, for HEAs, catalytic curves can attain many shapes since the distances between the current wave segments are directly correlated to the distances of the peak maxima within the AEDP (left-down of Fig. 17c). Replacing or adding element into the as-prepared HEAs can change all individual adsorption peaks, which has a direct influence on the position of the respective “current waves” (right of Fig. 17c). It is difficult to predict the real effect (positive or negative) of adding or replacing elements. However, the shape of the catalytic curve in the kinetic region provides direct information about the AEDP, that is, an integrated peak close to the optimal binding energy will yield a wave segment at low overpotentials with high currents. This represents an ideal result for the optimized HEA catalysts. In addition, they also developed noble metal-free CrMnFeCoNi nanoparticles with high intrinsic ORR activity [87]. Combining co-sputtering into an ionic liquid and potential-assisted immobilization of the formed nanoparticles at a microelectrode, the CrMnFeCoNi HEAs with a diameter of about (1.7 ± 0.2) nm yielded superior intrinsic activity that could compete with that of Pt.

Fig. 17

a Simplified scheme to illustrate active site distribution within one adsorption peak. Active sites of similar activity are grouped together and summarized as one site, whose intensity depends on the amount of similar sites; b schematic illustration of correlations between complex solid solution nanoparticle structures, its effect on the adsorption energy distribution pattern, and the respective electrochemical response in the kinetic region. The shift in energy of the individual adsorption peaks upon replacement or addition of elements depends on the material-inherent properties based on the complex interaction with the other elements and only few arbitrary examples of the many options are visualized. Reproduced with permission from Ref. [88]. Copyright 2019, Wiley–VCH

Recently, Hu et al. [134] fabricated a record-high carbon-supported HEOs nanoparticles composed of ten metal elements, i.e., Hf, Zr, La, V, Ce, Ti, Nd, Gd, Y, and Pd by the carbothermal shock strategy (Fig. 18c). Considering that conventional electrocatalysts suffer from performance degradation during long-term operations as a result of nanoparticles detachment and agglomeration (Fig. 18a), this stable electrocatalyst offers excellent structural, chemical as well as interfacial stability (Fig. 18b), enabled by the far-equilibrium synthetic approach and high-entropy design. Consequently, this 10-HEO nanoparticle on carbon black (10-HEO/C) presents both excellent ORR activity and significantly improved stability (92% and 86% retention after 12 and 100 h operation in an O₂-saturated 0.1 M KOH electrolyte by using a constant potential of 0.60 V vs. RHE and 1600 r min⁻¹, respectively) compared to the commercial Pd/C electrocatalyst (76% retention after 12 h operation in an O₂-saturated 0.1 M KOH electrolyte using a constant potential of 0.60 V vs. RHE and 1 600 r min⁻¹) (Fig. 18d). The outstanding electrocatalytic performance could be potentially benefited from the synergistic effect of 10 metal elements and higher utilization of active sites, and the high-entropy nature and the strong interfacial bonding will offer desired stability. In addition to HEAs nanoparticles, the multi-metallic nanosheets/nanoplates also display superior catalytic performance and long-term catalytic durability for the ORR compared with commercial Pt/C catalysts. By utilizing solvothermal strategy, Wang et al. [91] successfully synthesized nanoporous PtAgBiCo multi-metallic nanoplates for electrocatalyzing ORR in direct methanol fuel cells. The specific and mass activity of the PtAgBiCo nanoplates were 8 and 5 times greater than that of the commercial Pt/C catalyst, respectively. The better ORR performance of the PtAgBiCo catalyst could be ascribed to stable morphology and porous surface structure of the nanoplates that can offer more reaction sites for ORR. Note that, the incorporation of Bi into the alloy might also increase the tolerance to methanol poisoning. Additionally, porous PtCuBiMn nanosheets by a similar solvothermal strategy were also synthesized [90]. The specific activity and mass activity of a PtCuBiMn catalyst were 2.41 mA cm⁻² at 0.90 V and 0.69 A mg_pt⁻¹ at 0.90 V, which was ten and seven times higher than that of the generally used commercial Pt/C (0.24 mA cm², 0.1 A mg_pt⁻¹), respectively. Besides, PtCuBiMn catalysts also showed only a loss of 3.8% in mass activity at 0.90 V after 10 000 cycles of the durability test (the electrolyte, O₂-saturated 0.1 M HClO₄ solutions; scanning potential, 0.6 − 1.0 V vs. RHE; the scanning rate, 50 mV s⁻¹).

Fig. 18

a Performance degradation during long-term operations due to nanoparticle detachment and agglomeration for traditional carbon-supported electrocatalysts; b carbon-supported high-entropy oxide nanoparticles, exhibits good activity and excellent structural, chemical as well as interfacial stabilities, enabled by the far-equilibrium synthetic approach (about 1 400 K, and 1 s) and high-entropy design; c schematic of the high-entropy oxide nanoparticles dispersed on carbon black as ORR electrocatalysts; d chronoamperometric profiles (i.e., the percentage of current retention as a function of time) of the commercial Pd/C and the 10-HEO/C, tested in an O₂-saturated 0.1 M KOH electrolyte using a constant potential of 0.60 V vs. RHE and 1 600 r min⁻¹. Reproduced with permission from Ref. [134]. Copyright 2021, Wiley–VCH

5 Conclusion and Future Prospects

Due to the complexity and tunability of element composition, HEA materials have become a hotspot in the research field of electrocatalysis [135, 136]. In the electrocatalytic process, multi-metal alloying can significantly adjust the adsorption energy of reactant molecules and intermediates on the catalyst surface, and thus bring about the synergistic effect between the homogeneously distributed five or more elements, and enhance catalytic activity of HEA materials. In addition, it is widely reported that HEAs with distinct 3D morphology structure can not only provide high surface area, but also create a large number of active sites with a low coordination number on high index crystal planes.

In this review, the current state of HEA materials has been summarized with emphasis on their synthesis techniques, characterization methods as well as advances in researches and applications as novel electrocatalysts. Starting from the synthesis methods that have the capability to alloy immiscible elements into single phase solid solution and even immobilize them on suitable supports, both advantages and disadvantages of top-down and bottom-up strategies are elaborated in order to provide an ordinary guideline for HEAs preparation. Then the common characterization techniques of physicochemical properties and thermodynamic calculations of HEAs, are summarized. Finally, the actual applications of HEA materials for diverse electrocatalytic reactions, including HER, OER and ORR, have been demonstrated to evaluate their functional properties and put forward some reasonable guidance for electrocatalytic performance enhancement.

The preparation and electrocatalytic applications of HEA materials are both an opportunity and a challenge for researchers. Despite the significant achievements have been made in recent years, some issues still have to be addressed in future:

• (i) The novel synthetic method should be further developed to meet the requirements of the size uniformity, elements applicability, substrates applicability, composition and structure controllability simultaneously. In this regard, synergizing HEA materials with nanostructures and distinct morphology should be given more effort. Moreover, the facile large-scale methods to prepare HEA materials with simple procedures, mild reaction conditions and low cost need to be further explored to promote their practical applications on practical production process.

• (ii) Unraveling the detailed structure-composition-activity relationship that account for the intriguing catalytic behaviors of HEAs remains a great challenge. Some advanced in-situ characterization techniques should be well developed to unveil the catalytic mechanisms of HEA materials as well as their deactivation evolution processes.

• (iii) More accurate and simple theoretical models for theoretical calculations are further needed for fundamental understanding of HEA and new HEA catalyst design and fabrication. The calculations can serve as the auxiliary methods to verify the catalytic active sites, the electronic structures and the strain effect of HEAs and provide foundation for catalysts synthesis. It is also essential for time-saving research and development of HEAs materials.

• (iv) The dealloying strategy is a promising way to enhance the catalytic activity of HEA materials, but it is still at an early stage and much more detailed work is needed.