Three Strategies for the Design of Advanced High-Entropy Alloys

High-entropy alloys (HEAs) have recently become a vibrant field of study in the metallic materials area. In the early years, the design of HEAs was more of an exploratory nature. The selection of compositions was somewhat arbitrary, and there was typically no specific goal to be achieved in the design. Very recently, however, the development of HEAs has gradually entered a different stage. Unlike the early alloys, HEAs developed nowadays are usually designed to meet clear goals, and have carefully chosen components, deliberately introduced multiple phases, and tailored microstructures. These alloys are referred to as advanced HEAs. In this paper, the progress in advanced HEAs is briefly reviewed. The design strategies for these materials are examined and are classified into three categories. Representative works in each category are presented. Finally, important issues and future directions in the development of advanced HEAs are pointed out and discussed.


Introduction
The concept of high-entropy alloys (HEAs) was proposed by Yeh et al. in 2004 [1].It refers to a class of alloys that is composed of five or more principal elements, whose concentrations fall in the range of 5-35 at.% [1].Some researchers adopt a different definition based directly on entropy, which considers alloys with configurational entropies higher than 1.5R (where R is the gas constant) as HEAs [2,3].Due to their multi-principal-element nature, HEAs possess four unique core effects [4,5]: high entropy, sluggish diffusion [6], severe lattice distortion, and cocktail effects.
In the past decade, significant attention has been paid to these materials mainly due to two reasons.Firstly, the concept of HEAs brings a whole new perspective and unprecedented opportunities to metallic materials.The conventional view of the alloy world is based on dozens of individual alloy families such as Ti alloys, Cu alloys, etc.This is because conventionally there has to be a solvent element for each alloy.The baseless concept in HEAs frees metallurgists from this constraint and opens up a new dimension (instead of just a few new alloy systems) in alloy design.Therefore, the invention of HEA has been referred to as "a renaissance in physical metallurgy" [7].Secondly, many reported HEAs have exciting properties, for example, high strength and toughness, good wear/corrosion resistance, high structural thermal stability, good resistance to thermal softening, and special electric or magnetic properties.It is worth mentioning that these properties were typically obtained without careful optimization of composition and/or microstructure.Therefore, it is anticipated that the properties of these materials can be further improved once related knowledge is developed.
Probably owing to the compositional complexity and the intimidating amount of possible element combinations, initial studies on HEAs largely focused on HEAs composed of common metallic elements, particularly Al, Co, Cr, Cu, Fe, and Ni.This leads to the two classic alloys systems that have been most extensively studied: Al-Co-Cr-Cu-Fe-Ni [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] and Al-Co-Cr-Fe-Ni [23][24][25][26][27][28].The crystal structure, phase transformation, microstructure, mechanical properties of these systems have been reported in detail, and the evolution of the above issues with respect to the content of each element have also been investigated.Considerable effort has also been devoted to the derivative systems of these classic systems.A system is said to be "derived" from another if the number of different principal element between the two is two or less [5].Researchers usually either replace one of the elements with a different element, or simply add other element into the classic systems.For example, derivative systems alloyed with Mo, Mn, Si, B, Zr, Ni, Nb, V, Y, Sn, Zn, C [29][30][31][32][33][34][35][36][37][38] have been reported.These early results provided important understanding of the composition-phase-property relationship in HEAs, which made composition design and phase control possible to certain extent.Despite the important contributions of the early studies, these studies are largely of an exploratory nature.For example, the composing elements of the early alloys are usually very similar.Although there can be one or two uncommon alloying elements, they are selected more or less arbitrarily.Additionally, the ratio of the elements are usually either equimolar or mostly equimolar (only one of the elements have varying content).Most importantly, there is no specific goal to be achieved in the design of the early alloys.
Very recently, the development of HEAs has gradually entered a different stage.Unlike the early alloys, HEAs studied nowadays are usually designed to meet clear goals, and have carefully chosen composing elements, deliberately introduced multiple phases, and tailored microstructures.These alloys will henceforth be referred to as advanced HEAs.In this paper, the design strategies of advanced HEAs are examined, and they can be classified into three categories.The progress within each category is then reviewed.The three categories are: (1) application-based redesign; (2) enhancing the entropy of conventional alloys; and (3) incorporating second phases.
It should be noted that the three strategies are not mutually exclusive.In other words, an alloy can be designed using more than one of the above strategies at the same time.This will be further explained later.It should be also noted that some of the alloy systems discussed in this paper (e.g., some high-entropy steels and brasses) are not HEAs by definition.However, because they are directly inspired by the high-entropy concept, these alloys are also discussed in the present paper.

Application-Based Redesigning
Application-based redesign is probably the most prevalent strategy of the three.It aims to design HEAs for a particular application.The selection of component elements is based specifically on the considerations of the desired properties, and could be very similar to those in existing conventional alloys for that purpose.Examples include refractory HEAs, light HEAs, HE bulk metallic glass (HEBMG), and high-entropy superalloys.

Refractory HEAs
First proposed by Senkov et al. in 2010, these are among the earliest examples of application-based redesigned materials [39].These alloys are composed of refractory elements and are aimed for high temperature structural applications [40][41][42][43][44][45][46][47].The primary phases in these alloys are generally BCC phase.Some alloys even have a single BCC structure.However, these multi-principal-element BCC phases are way stronger than pure BCC metals.For example, the hardness of the Mo 20 Nb 20 Ta 20 V 20 W 20 alloy is 5250 MPa [39].More importantly, these alloys can be very resistant to thermal softening and can retain high hardness to temperatures as high as 1600 ˝C [42].This is attributed to the sluggish diffusion effect in HEAs [42,48].Therefore, these alloys have great potential in high temperature applications.Many other refractory HEAs were developed later to obtain improved strength, room temperature ductility, and density [40,41,[47][48][49][50][51][52][53][54].For example, Al addition in certain systems can bring advantages such as higher hardness and high-temperature strength, better room temperature plasticity, and lower density [48].Additionally, NbMoTaW pillars show extraordinarily high yield strength of nearly 10 GPa.This is among the highest reported strength in micro-/nano-pillar compression.Furthermore, its fracture strain under compression is as high as 30% [55].

Light HEAs
Because of the strong demand for energy conservation, the development of advanced light metals has been an important issue.However, conventional light metals often have mediocre strength and limited high temperature applications.In contrast, HEAs are known for high strength, good resistance to thermal softening, and good structural thermal stability [5].Thus, light HEAs is a direction with great potential.Candidate elements include all low-density elements (e.g., Al, Mg, Ti, Si, V, Li, Ca, Sc, . . . ) and other elements to optimize properties.For example, Juan et Li 20 Mg 10 Sc 20 Ti 30 , which also has very low density (2.67 g/cm 3 ) [59].According to these studies, light HEAs do have high strength.As long as reasonable ductility can be achieved, these alloys will have great potential in the industry.

High-Entropy Bulk Metallic Glass (HEBMG)
Bulk metallic glasses (BMGs) have very high strength and good corrosion resistance.Thus, they have been extensively studied in the past few decades [60][61][62][63][64].The composition of BMG is similar to HEAs in that both materials have multiple components.However, HEAs do not have a solvent element, and the configurational entropy is higher.It has been pointed out that higher configuration entropy could be beneficial for the glass forming ability (GFA) of BMGs [65,66].Thus, it is reasonable to expect that some HEAs can be good glass formers.However, since most reported HEAs are crystalline, high configurational entropy apparently does not directly translate to glass formation.Other factors also play critical roles: Guo et al. showed that HEAs do form metallic glasses when they have large negative enthalpy of mixing and large atomic size difference [67].These conditions are in line with the empirical rules for BMG formation [64].The number of reported high entropy BMGs is still not large [67][68][69][70][71][72][73][74][75].Some of the reported alloys do have larger critical diameters.For example, the diameters for Pd 20 Pt 20 Cu 20 Ni 20 P 20 and Ti 20 Zr 20 Hf 20 Be 20 Ni 20 glasses are 1 and 3 mm, respectively [76,77].The latter has a yield strength of 2142 MPa and a plasticity of 4% [77].HEBMGs can have other interesting properties.For example, Zn 20 Ca 20 Sr 20 Yb 20 (Li 0.55 Mg 0.45 ) 20 shows homogeneous flow and a remarkable plasticity (~25%) at room temperature [68].CaMgZnSrYb, a high-entropy modification of Ca 65 Mg 15 Zn 20 BMG, was found to promote osteogenesis and bone formation after two weeks of implantation [69].High entropy metallic glasses could also have improved thermal stability.Amorphous NbSiTaTiZr alloy film has an exceptionally high crystallization temperature of over 800 ˝C, which is probably the highest in any reported metallic glasses [78].

High-Entropy Superalloys (HESA)
High-entropy superalloys were proposed in 2015 [79].Conventionally, the most important family of superalloys are the Ni-based ones.The replacement of Ni by other elements could reduce cost and/or density.The high structural stability and slow diffusion kinetics might be helpful for the thermal stability of the gamma prime phase.Yeh et al. developed a series of HESAs based on the Al-Co-Cr-Fe-Ni-Ti system [79][80][81][82].These alloys do have reduced densities (below 8 g/cm 3 ) and lower raw material costs (about 20% lower than common Ni-based superalloys).Among these alloys, Al 0.1 Co 1.5 CrFeNi 1.5 Ti 0.4 has good oxidation resistance, high hot hardness (~350 Hv at 800 ˝C), and does not contain the detrimental η phase [81].Ni 44 Al 3.9 Co 22.3 Cr 11.7 Fe 11.8 Ti 6.3 (in wt.%) and Ni 51 Al 5 Co 18 Cr 7 Fe 9 Ti 5 Ta 2 Mo 1.5 W 1.5 (in wt.%) show exceptional oxidation resistance and hot corrosion resistance that is superior to commercial superalloy CM247LC (Figure 1) [82].Mazoni et al. optimized the Al-Co-Cr-Cu-Fe-Ni-Ti system [83][84][85][86] and obtained a promising alloy Al 10 Co 25 Cr 8 Fe 15 Ni 36 Ti 6 , which shows the γ-γ 1 microstructure characteristic of superalloys.This alloy has a tensile strength of 650 MPa at 800 ˝C, which is higher than that of Inconel 617 (less than 500 MPa) [85].Senkov et al. used a very different approach to obtain HESAs [54].Using mostly refractory elements, they synthesized an alloy AlMo 0.5 NbTa 0.5 TiZr that contains high density cubic/plate-like precipitates embedded in a matrix-which looks similar to the structures of Ni-based superalloys.The precipitates have a disordered BCC structure, while the matrix has an ordered B2 structure.The strength of this alloy is significantly higher than Ni-based superalloys in the temperature range of 20 ˝C to 1200 ˝C [54].Besides the above examples, there are many more other cases in this category.High-entropy hydrogen storage material is also an appealing direction.The CoFeMnTiVZr alloy can absorb/desorb hydrogen up to 1.6 wt.% at room temperature [87].The ZrTiVCrFeNi alloy has a maximum hydrogen capacity of 1.8 wt.% when operated at 50 °C [88].High-entropy magnetic materials also have interesting properties.For example, the FeCoNiAl0.2Si0.2alloy has a good combination of properties including high saturation magnetization (1.151 T), high resistivity (69.5 μΩ-cm), and good malleability [89].High-entropy diffusion barriers [78,[90][91][92], high-entropy thermoelectric materials [93], high entropy protective coatings [94][95][96][97][98], etc., have also been reported.In fact, because materials for almost every application can be redesigned with the high-entropy concept, it is expected that there will be significantly more alloy systems in this category in the future.

Enhancing the Entropy of Conventional Alloys
The second strategy is to modify the composition of a known conventional alloy using the high-entropy concept.The modification can be made either to the solvent or to the solutes, and is illustrated in Figure 2a.The key point is to replace the existing solvent/solute with more elements.Modification of the solvent usually leads directly to a new HEA.In contrast, modification of the solute leads to a conventional alloy with enhanced entropy.
An early example of entropy enhancement of the solvent is in the field of BMG [99,100].Based on the Zr-Al-Ni alloy, Cantor et al. performed what they called equiatomic substitution and developed the (Ti33Zr33Hf33)100−x−y(Ni50Cu50)xAly alloys [101].The original solvent, Zr, is replaced by Ti, Zr, Hf in equiatomic ratio.Among these alloys, (Ti33Zr33Hf33)60(Ni50Cu50)20Al20 showed very good GFA, with a wide supercooled liquid region of 124 K [102].Another example is high-entropy bronzes/brasses [103].These novel materials are developed by noting that Cu-Mn-Ni are mutually soluble.Therefore, the authors replaced the Cu element in conventional bronzes and brasses with equiatomic Cu-Mn-Ni ternary alloy, leading to compositions like (CuMnNi)95Sn5 or (CuMnNi)80Zn20.The strength, hardness, and plasticity of these alloys are considerably higher than standard commercial bronzes and brasses [103].Besides the above examples, there are many more other cases in this category.High-entropy hydrogen storage material is also an appealing direction.The CoFeMnTiVZr alloy can absorb/desorb hydrogen up to 1.6 wt.% at room temperature [87].The ZrTiVCrFeNi alloy has a maximum hydrogen capacity of 1.8 wt.% when operated at 50 ˝C [88].High-entropy magnetic materials also have interesting properties.For example, the FeCoNiAl 0.2 Si 0.2 alloy has a good combination of properties including high saturation magnetization (1.151 T), high resistivity (69.5 µΩ-cm), and good malleability [89].High-entropy diffusion barriers [78,[90][91][92], high-entropy thermoelectric materials [93], high entropy protective coatings [94][95][96][97][98], etc., have also been reported.In fact, because materials for almost every application can be redesigned with the high-entropy concept, it is expected that there will be significantly more alloy systems in this category in the future.

Enhancing the Entropy of Conventional Alloys
The second strategy is to modify the composition of a known conventional alloy using the high-entropy concept.The modification can be made either to the solvent or to the solutes, and is illustrated in Figure 2a.The key point is to replace the existing solvent/solute with more elements.Modification of the solvent usually leads directly to a new HEA.In contrast, modification of the solute leads to a conventional alloy with enhanced entropy.
An x Al y alloys [101].The original solvent, Zr, is replaced by Ti, Zr, Hf in equiatomic ratio.Among these alloys, (Ti 33 Zr 33 Hf 33 ) 60 (Ni 50 Cu 50 ) 20 Al 20 showed very good GFA, with a wide supercooled liquid region of 124 K [102].Another example is high-entropy bronzes/brasses [103].These novel materials are developed by noting that Cu-Mn-Ni are mutually soluble.Therefore, the authors replaced the Cu element in conventional bronzes and brasses with equiatomic Cu-Mn-Ni ternary alloy, leading to compositions like (CuMnNi) 95 Sn 5 or (CuMnNi) 80 Zn 20 .The strength, hardness, and plasticity of these alloys are considerably higher than standard commercial bronzes and brasses [103].Entropy enhancement of the solute has been conducted to develop various high-entropy steels and high-entropy cast iron.Raabe et al. designed high entropy steels based on the Fe-Mn-Al-Si-C system [104].They introduced more solutes to the Fe-Mn-C system and used the increased configurational entropy to stabilize a single-phase homogeneous FCC phase.Their results are quite pleasing-ultimate tensile strengths of up to 1000 MPa and tensile elongations reaching up to 100% were reported [104].Compared to conventional steels, these high-entropy steels possess a remarkable combination of properties, which is shown in Figure 3.
A very different yet also interesting type of high-entropy steel is introduced by Pushin et al. [105].They alloyed many strong carbide-forming elements (Cr, Mo, W, and V) to the Fe-C system at the same time.The result is an ultra-fine-dispersed composite martensite-multicarbide structure with a very high hardness of HRC 62.It was noted that both binary and multi-component carbides were observed.A similar modification was made by Wang et al. to white cast iron Fe-20Cr-5C in 2011 [106].They alloyed several carbide-forming elements such as Ti, V, Mo, and W simultaneously.The alloying led to two effects.Firstly, it suppressed the formation of primary coarse carbides and instead promoted the formation of various new carbide types (M7C3, MC, and M6C).Secondly, it significantly refined the microstructure of the alloy owing to the competition between the carbide-forming elements.Therefore, the wear resistances of the high-entropy cast irons were markedly improved.For example, the volume loss rate of the 54Fe15.46Cr16.76C-3.45TiVMoWhigh-entropy cast iron is only one thirds of the original white cast iron.Based on the above examples, one can see that the entropy-enhanced alloys do show improved properties over the original ones.It is also noted that the composition, phase, and microstructure of the modified alloys can be very different.Depending on the modification made, the modified alloy could be an HEA or a conventional alloy, single-phased or multi-phased.In the case of a multi-phased alloy, the second phase(s) could be one high-entropy phase, various individual near-binary phases, or a mixture of the two extremes.These possibilities are partially illustrated with Figure 2b.The wide diversity in the structure and phase of the modified alloys allows considerable freedom for the adjustment of properties, making this strategy useful for the improvement of conventional alloy systems.Entropy enhancement of the solute has been conducted to develop various high-entropy steels and high-entropy cast iron.Raabe et al. designed high entropy steels based on the Fe-Mn-Al-Si-C system [104].They introduced more solutes to the Fe-Mn-C system and used the increased configurational entropy to stabilize a single-phase homogeneous FCC phase.Their results are quite pleasing-ultimate tensile strengths of up to 1000 MPa and tensile elongations reaching up to 100% were reported [104].Compared to conventional steels, these high-entropy steels possess a remarkable combination of properties, which is shown in Figure 3.

Incorporating Second Phases
HEAs are known to form simple phases (i.e., FCC, BCC, and HCP structures, including both ordered and disordered versions) easily, and these simple phases can have many unique properties [55,[107][108][109][110].However, the strengths of simple-phased HEAs are usually limited.In A very different yet also interesting type of high-entropy steel is introduced by Pushin et al. [105].They alloyed many strong carbide-forming elements (Cr, Mo, W, and V) to the Fe-C system at the same time.The result is an ultra-fine-dispersed composite martensite-multicarbide structure with a very high hardness of HRC 62.It was noted that both binary and multi-component carbides were observed.A similar modification was made by Wang et al. to white cast iron Fe-20Cr-5C in 2011 [106].They alloyed several carbide-forming elements such as Ti, V, Mo, and W simultaneously.The alloying led to two effects.Firstly, it suppressed the formation of primary coarse carbides and instead promoted the formation of various new carbide types (M 7 C 3 , MC, and M 6 C).Secondly, it significantly refined the microstructure of the alloy owing to the competition between the carbide-forming elements.Therefore, the wear resistances of the high-entropy cast irons were markedly improved.For example, the volume loss rate of the 54Fe15.46Cr16.76C-3.45TiVMoWhigh-entropy cast iron is only one thirds of the original white cast iron.Based on the above examples, one can see that the entropy-enhanced alloys do show improved properties over the original ones.It is also noted that the composition, phase, and microstructure of the modified alloys can be very different.Depending on the modification made, the modified alloy could be an HEA or a conventional alloy, single-phased or multi-phased.In the case of a multi-phased alloy, the second phase(s) could be one high-entropy phase, various individual near-binary phases, or a mixture of the two extremes.These possibilities are partially illustrated with Figure 2b.The wide diversity in the structure and phase of the modified alloys allows considerable freedom for the adjustment of properties, making this strategy useful for the improvement of conventional alloy systems.

Incorporating Second Phases
HEAs are known to form simple phases (i.e., FCC, BCC, and HCP structures, including both ordered and disordered versions) easily, and these simple phases can have many unique properties [55,[107][108][109][110].However, the strengths of simple-phased HEAs are usually limited.In particular, alloys that have FCC structures are quite soft.Among common strengthening mechanisms, second-phase strengthening is known to be the most effective [3,111].Second phases can be incorporated directly in the cast state, or be precipitated in a matrix after heat treatment.HEAs have slow diffusion kinetics [6] and can form nanoscale second phase/precipitates in the as-cast state [9,10,23,112], after prolonged annealing at 700 ˝C [113], and even after homogenization at 1100 ˝C and subsequently furnace cooled [11].Therefore, HEAs are very suitable for this strengthening mechanism.Indeed, it has been reported that second phases can markedly enhance the mechanical properties of HEAs.For example, the precipitation of hard η-Ni 3 Ti phase drastically improve the wear resistance of the Co 27.3 Cr 18.2 Fe 18.2 Ni 27.3 Ti 9.1 alloy [114].Another example is in the AlCoCr x FeMo 0.5 Ni alloy, where the formation of σ phase leads to a better resistance to thermal softening at elevated temperatures [115].These results reveal that second-phase strengthening works well in HEAs.
In view of this, very recently many groups aim to design HEAs with intentionally introduced second phase with tailored morphology.For example, He et al. alloyed a small amount of Ti and Al into FeCoNiCr to precipitate nanosized coherent reinforcing phase, L1 2 -Ni 3 (Ti, Al), in an FCC FeCoNiCr matrix [116].This γ-γ 1 structure is similar to that in Ni-based superalloys and leads to significant precipitation hardening effect.Thermomechanically processed (FeCoNiCr) 94 Ti 2 Al 4 alloy can reach yield strength of over 1 GPa while still remain of good elongation of 17%.Such combination of properties outperforms most advanced steels, as is shown in Figure 4 [116].Guo et al., on the other hand, alloyed C into the Mo 0.5 NbHf 0.5 ZrTi refractory HEA [117].The alloying leads to the formation of MC (M is Hf, Nb, Zr, and Ti) carbide particulates.The Mo 0.5 NbHf 0.5 ZrTiC 0.1 alloy shows a compressive strength of 2139 MPa and a fracture strain of 38.4%.Compared with the corresponding values of the original alloy (1538 MPa and 24.6%), addition of C improves both the strength and plasticity of the alloy.Another example is AlCoCrFeNi 2.1 , an alloy which comprises two phases: FCC and B2.In its cast state, the two phases form a lamellar structure and the alloy is soft.It can be thermomechanically processed to achieve an ultrafine grained duplex microstructure which has well-balanced mechanical properties-a yield strength of 1100 MPa and an elongation of 12% [118].
Entropy 2016, 18, 252; doi:10.3390/e18070252www.mdpi.com/journal/entropyBased on the above, the introduction of appropriate second phases with tailored morphology is a very effective strategy for property enhancement.However, the key difficulty in this regard lies in the complexity of composition design.This will be discussed further in the final part of this paper.[116].A: FeCoNiCr, as-homogenized; B: (FeCoNiCr)94Ti2Al4, as-homogenized; P1: FeCoNiCr)94Ti2Al4, thermomechanically processed; and P2: FeCoNiCr)94Ti2Al4, thermomechanically processed.The thermomechanical processing of P1 and P2 are different, see the original work for details.

Notes on the Classification of the Alloys
Some notes on the classification of the alloy are presented here.Firstly, this paper discusses the design strategy of advanced HEAs.Therefore, the classification of alloys is based on the design concept described in the original publication, not on the phase/structure of the resulted alloy.For example, carbide strengthened high-entropy steel is classified in the second category instead of the third because the design concept of the author is based on enhancing the entropy of an existing steel.
Secondly, the three strategies are not mutually exclusive; one can adopt one or more strategies simultaneously to design an alloy.For example, HE superalloys are designed to replace current Ni-based superalloys; which means they are designed specifically for high temperature structural applications [80].For that purpose, the gamma/gamma prime microstructure in Ni-based superalloys is known to be very helpful.Thus, the authors also want to mimic that microstructure.Therefore, both the first and the third strategies are used in this case.
Thirdly, when more than one strategy is used to design an alloy, the alloy should be classified under the category with higher hierarchy whenever possible.For example, in the case of HE superalloys, the incorporation of the gamma prime phase is because it benefits the high temperature strength.In other words, the third strategy is adopted because of the need in the first strategy.Therefore, HE superalloys are classified under the first category.

Outlook and Concluding Remarks
Some important issues in the future development of advanced HEAs are discussed below.Firstly, the composition design of HEAs is still challenging and sometimes intimidating.This is particularly true when intermetallic phases are involved.Our current knowledge about the design principles of HEAs is largely focused on designing simple-structured HEAs (i.e., FCC, BCC, HCP, and their derivative structures) [5,110].However, because many advanced HEAs contain multiple phases including intermetallics, knowledge about controlling the type and amount of intermetallic phases becomes crucial.In this regard, the following directions are worth studying: (1) High-entropy intermetallics [119].Although in many cases the intermetallics in HEAs are just conventional binary/ternary intermetallics, sometimes the so-called high-entropy intermetallics (HEI) can appear [119].HEIs have crystal structures similar to binary intermetallic phases but they are composed of multiple principal elements.For example, the composition of the Laves Based on the above, the introduction of appropriate second phases with tailored morphology is a very effective strategy for property enhancement.However, the key difficulty in this regard lies in the complexity of composition design.This will be discussed further in the final part of this paper.

Notes on the Classification of the Alloys
Some notes on the classification of the alloy are presented here.Firstly, this paper discusses the design strategy of advanced HEAs.Therefore, the classification of alloys is based on the design concept described in the original publication, not on the phase/structure of the resulted alloy.For example, carbide strengthened high-entropy steel is classified in the second category instead of the third because the design concept of the author is based on enhancing the entropy of an existing steel.
Secondly, the three strategies are not mutually exclusive; one can adopt one or more strategies simultaneously to design an alloy.For example, HE superalloys are designed to replace current Ni-based superalloys; which means they are designed specifically for high temperature structural applications [80].For that purpose, the gamma/gamma prime microstructure in Ni-based superalloys is known to be very helpful.Thus, the authors also want to mimic that microstructure.Therefore, both the first and the third strategies are used in this case.
Thirdly, when more than one strategy is used to design an alloy, the alloy should be classified under the category with higher hierarchy whenever possible.For example, in the case of HE superalloys, the incorporation of the gamma prime phase is because it benefits the high temperature strength.In other words, the third strategy is adopted because of the need in the first strategy.Therefore, HE superalloys are classified under the first category.

Outlook and Concluding Remarks
Some important issues in the future development of advanced HEAs are discussed below.Firstly, the composition design of HEAs is still challenging and sometimes intimidating.This is particularly true when intermetallic phases are involved.Our current knowledge about the design principles of HEAs is largely focused on designing simple-structured HEAs (i.e., FCC, BCC, HCP, and their derivative structures) [5,110].However, because many advanced HEAs contain multiple phases including intermetallics, knowledge about controlling the type and amount of intermetallic phases becomes crucial.In this regard, the following directions are worth studying: (1) High-entropy intermetallics.Although in many cases the intermetallics in HEAs are just conventional binary/ternary intermetallics, sometimes the so-called high-entropy intermetallics (HEI) can appear.HEIs have crystal structures similar to binary intermetallic phases but they are composed of multiple principal elements.For example, the composition of the Laves phase in CoCrFeNiHf is (in at.%): 15.1% Cr, 19.4% Fe, 21.4% Co, 20.1% Ni, and 24.0%Hf [119,120].This composition is not far from a quinary equimolar alloy, which is apparently different from conventional binary/ternary Laves phases.High-entropy intermetallics represent a new class of material.Their behavior and properties could be different from their conventional counterparts and deserve more attention.(2) CALPHAD technique.The compositional space in HEAs is simply too enormous to explore by experiments alone.This makes computational skills, in particular CALPHAD, extremely important.It was suggested [3] and later on verified that CALPHAD can quickly screen out alloys that are apparently unsuitable, thus the speed of development can be increased by several orders of magnitude [121][122][123].A major issue is that the thermodynamic database for HEAs is far from complete.This means that depending on alloy composition, the accuracy of CALPHAD predictions can vary significantly.Therefore, the development of more multi-component databases is urgently needed.(3) Practical design rules concerning intermetallic phases in HEAs.Although the CALPHAD approach has unparalleled advantage in term of speed, the software and databases are costly and accuracy is still a major issue (as mentioned above).Therefore, practical alloy design principles based on experimental results is still necessary.Unfortunately, our knowledge about the stability and selection of intermetallic phases in HEAs is very rare.Preliminary results includes a general criterion on the selection of intermetallic phase [120], and criteria specific to certain phases types such as sigma phase [124,125] and Laves phase [126].Clearly, more work is required in this respect.
Secondly, because many advanced HEAs are aimed for real applications, cost becomes a key issue.This is a major drawback of HEAs.The introduction of multiple principal elements makes it difficult to use only cheap elements.Therefore, the following directions are of interest: (1) Focus on replacing expensive conventional alloys.A reasonable approach is to find applications where the current material is already expensive.Then, use the three strategies mentioned in this paper to find better solutions.Since the original material is expensive, the barrier of using HEAs is lowered as long as HEAs provide improved properties.Ni-based superalloys and Zr alloys are good examples of expensive conventional alloys.(2) Avoid expensive elements.The use of cheap elements directly reduces the material cost.For example, Co is an expensive element but it is very commonly used in HEAs.Therefore, alloys with reduced or no Co have clear advantage.Indeed, some researchers have explored Co-free HEAs [127,128].Most of these studies start from a known HEA with good properties and try to replace Co with other elements.The difficult part, of course, is to preserve the good properties of the original alloy.Co is not the only expensive element used in HEAs.It is expected that the replacement of expensive elements will become a common practice in alloy design.(3) Explore medium-entropy alloys.Medium-entropy alloys (MEAs) have been highlighted as an important future direction because of their great potential [129,130].But MEAs have another competitive edge.The reduced alloying content in MEAs can lower the material cost, particularly if the base element is also cheap.Fe-based MEAs are good examples.In this regard there is still a lot of space to explore.
In summary, the development of HEAs has entered a different stage in recent years.Unlike early studies that are largely of exploratory nature, HEAs studied nowadays are usually designed to meet clear goals, and have carefully chosen composing elements, deliberately introduced multiple phases, and tailored microstructures.These alloys are referred to here as advanced HEAs.The migration to advanced HEAs clearly signifies the progress of our knowledge on these novel materials.Based on a broad literature survey on advanced HEAs, the design strategies of these materials are classified into three categories, namely application-based redesigning, enhancing the entropy of conventional alloys, and incorporating second phases.Representative works in each category are discussed.It should be noted that the three strategies are not mutually exclusive.Finally, important issues in the development of advanced HEAs are pointed out.These include difficulties in alloy design and cost reduction.Future directions are also discussed accordingly.
early example of entropy enhancement of the solvent is in the field of BMG [99,100].Based on the Zr-Al-Ni alloy, Cantor et al. performed what they called equiatomic substitution and developed the (Ti 33 Zr 33 Hf 33 ) 100´x´y (Ni 50 Cu 50 )

Figure 2 .
Figure 2. (a) Two approaches to enhance the entropy of conventional alloys.Modification of solvent typically leads to a new HEA, while modification of solute leads to an entropy-enhanced conventional alloy.Each color represents one element and the width of the colored region represents its concentration; (b) Three possible ideal cases of phase content in the alloy after solute modification.(I) single-phased; (II) one high-entropy second phase; (III) several near-binary second phases.Each color represents one phase in the alloy and the width of the phase region represents its volume fraction.

Figure 2 .
Figure 2. (a) Two approaches to enhance the entropy of conventional alloys.Modification of solvent typically leads to a new HEA, while modification of solute leads to an entropy-enhanced conventional alloy.Each color represents one element and the width of the colored region represents its concentration; (b) Three possible ideal cases of phase content in the alloy after solute modification.(I) single-phased; (II) one high-entropy second phase; (III) several near-binary second phases.Each color represents one phase in the alloy and the width of the phase region represents its volume fraction.

Entropy 2016, 18 , 252 6 of 14 Figure 3 .
Figure 3. Overview diagram showing typical ranges of total elongation to fracture and ultimate tensile strength for a number of different classes of steels.Mechanical properties of reported high-entropy steels exceed that of many known steels [104].

Figure 3 .
Figure 3. Overview diagram showing typical ranges of total elongation to fracture and ultimate tensile strength for a number of different classes of steels.Mechanical properties of reported high-entropy steels exceed that of many known steels [104].