You are on page 1of 10

ARTICLE IN PRESS

JMADE-02515; No of Pages 10
Materials and Design xxx (2016) xxxxxx

Contents lists available at ScienceDirect

Materials and Design


journal homepage: www.elsevier.com/locate/matdes

Alloying effect on the elastic properties of refractory high-entropy alloys


Li-Yun Tian a, b , Guisheng Wang c , Joshua S. Harris d , Douglas L. Irving d ,
Jijun Zhao b , Levente Vitos a, e, f
a

Applied Materials Physics, Department of Materials Science and Engineering, Royal Institute of Technology, Stockholm SE-100 44, Sweden
Key Laboratory of Materials Modication by Laser, Ion, and Electron Beams of Ministry of Education, Dalian University of Technology, Dalian 116024, China
c
School of Applied Mathematics and Physics, Beijing University of Technology, Beijing100124, China
d
Department of Materials Science and Engineering, North Carolina State University, Raleigh, NC27695-7907, United States
e
Department of Physics and Astronomy, Division of Materials Theory, Uppsala University, Box 516, SE-75121 Uppsala, Sweden
f
Research Institute for Solid State Physics and Optics, Wigner Research Center for Physics, P.O. Box 49, Budapest H-1525, Hungary
b

A R T I C L E

I N F O

Article history:
Received 2 October 2016
Received in revised form 3 November 2016
Accepted 21 November 2016
Available online xxxx
Keywords:
High-entropy alloys
Lattice parameter
Elastic constant
Local lattice distortion
Alloying effect

A B S T R A C T
Ab initio total energy calculations are used to determine the elastic properties of TiZrVNb, TiZrNbMo and
TiZrVNbMo high-entropy alloys in the body centered cubic (bcc) crystallographic phase. Calculations are
performed using the Vienna Ab initio Simulation Package and the Exact Mun-Tin Orbitals methods, and the
compositional disorder is treated within the frameworks of the special quasi-random structures technique
and the coherent potential approximation, respectively. Special emphasis is given to the effect of local lattice
distortion and trends against composition. Signicant distortion can be observed in the relaxed cells, which
result in an overlap of the rst and second nearest neighbor (NN) shells represented in the histograms.
When going from the four-component alloys TiZrVNb and TiZrNbMo to the ve-component TiZrVNbMo,
the changes in the elastic parameters follow the expected trends, except that of C44 which decreases upon
adding equiatomic Mo to TiZrVNb despite of the large shear elastic constant of elemental Mo. Although the
rule of mixtures turns out to be a useful tool to estimate the elastic properties of the present HEAs, to capture
the more delicate alloying effects one needs to resort to ab initio results.
2016 Published by Elsevier Ltd.

1. Introduction
High-entropy alloys (HEAs) are equiatomic or nearly-equiatomic
multicomponent solid solutions, which attracted substantial attentions in recent years [16]. They crystalize in body centered cubic
(bcc) [713], face centered cubic [1416] or hexagonal close-packed
(hcp) lattice [17], and often show a good combination of high ductility and high strength. Due to the rich physical, chemical and
mechanical properties [9,14,18], HEAs are considered as the primary choice in many future high-technology applications. The large
number of possible combinations of ve or more metallic elements
makes the number of potential HEAs enormous [19,20]. Despite of
the substantial experimental [21,22] and most recently also theoretical [19,23-30] efforts to explore the HEAs, today the available
information on their properties is rather limited.
It is often assumed that the bulk parameters of random solid
solutions may be estimated from the rule of mixtures [3133]
representing a weighted mean of the parameters of pure constituents. This is especially the case for non-magnetic alloys consisting of elements which in their pure form have the same crystal
lattice. A well-known example is Vegards law [34], which is used

to evaluate the equilibrium lattice parameter of alloys using data


available for the end members. Such simple rules are very useful
for an initial screening of the alloying effects for unknown systems such as new HEAs and may be used for categorizing the alloy
components for design purposes. In theoretical condensed matter
physics, subtle electronic structure mechanisms are traced which are
responsible for the small deviations relative to these rules. Unfortunately, today the available experimental data on HEAs has no
sucient volume and accuracy to provide a solid ground for assessing the rule of mixtures for these multi-component alloys. Some
initial attempts of using rst principles approaches to evaluate simple mixing rules have been done for composition variations NiFeCrCo [35] and MoNbTaVW [36] but little has been done for refractory
systems.
The TiHfZrTaNb refractory high entropy alloy was found to
show superior mechanical properties [37]. The corresponding elastic
properties were also investigated by ultrasound measurements
yielding C44 = 28 GPa and C11 = 172 GPa, Young modulus E =
78.5 GPa, the bulk modulus B = 134.6 GPa and Poisson ratio l =
0.402 [38]. Unfortunately, experimental data on single-crystal or
polycrystalline elastic moduli for other refractory HEAs are very

http://dx.doi.org/10.1016/j.matdes.2016.11.079
0264-1275/ 2016 Published by Elsevier Ltd.

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
2

L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

limited, making all experimental and theoretical attempts to derive


and understand such data highly timely and welcome.
First-principles methods built on density functional theory
(DFT) [3941] have been widely used to access the properties of
solids including multi-component disordered alloys. Special quasirandom structure (SQS) [42,43] and coherent potential approximation (CPA) [4446] are two frequently used techniques in combination with DFT methods for modeling the random distribution of
alloy components on a given crystal lattice. SQS involves large supercells constructed in terms of atomic correlation functions mimicking
a random distribution within the rst few coordination shells. The
atomic positions within the supercells may be relaxed by minimizing
the interatomic forces during the self-consistent calculations and by
that properly accounting for the local lattice distortion (LLD) effect.
Hence, SQS combined with a suitable underlying DFT method may
provide highly accurate results for solid solutions. The main limiting factor of this technique is the computational load associated with
large supercells needed to model (nearly) arbitrary compositions.
In addition, these supercells usually have lower symmetry than the
underlying lattice which further increases the computational efforts.
The CPA is a very powerful mean-eld technique to treat disordered systems. The real alloy is substituted by an effective medium
obtained by solving the impurity problem for each alloy constituent
within the single-site approximation. This approach is the ideal
choice for complex multi-component alloys such as HEAs since the
size of the fundamental DFT equations scales linearly with the number of alloy components. In addition, the symmetry of the alloy
matrix is retained which facilitates the calculations of symmetry
dependent properties such as elastic constants. The main drawback
of CPA is that the LLD can not be taken into account within the singlesite approximation. Because of that, most theoretical research on
HEAs using CPA focused on the alloying effect but completely omitted the LLD. On the other hand, severe lattice distortion decreases the
XRD peak intensity [47] and contributes signicantly to properties
of HEAs. Since nowadays CPA is commonly used in the description
of HEAs, it is highly desired to establish the impact of LLD on the
properties in question and check whether the associated errors can
overwrite the chemical effects.
High entropy alloys containing refractory elements (Ti, Zr, Hf, Nb,
Mo, W, V and Ta) display a single-phase bcc structure [7,8]. These
alloys exhibit superior mechanical properties in compression at high
temperature compared to the conventional high-temperature alloys.
Experimentally, Senkov et al. showed that the NbTiVZr alloy adopts a
single bcc phase [48]. The XRD patterns of TiZrNbMoVx (x = 0, 0.25,
0.5, 0.75, 1) show only a single bcc phase and those of TiZrNbMoVx
(x = 1.5, 2, 3) exhibit both bcc and Zr-rich (b-Zr) phases with
increasing V molar ratio [22,49]. These alloys have been in the
focus of many experimental and theoretical studies [4,9,23,48,50].
Dirras et al. investigated the elastic properties of TiHfZrTaNb using
ultrasound measurements [38]. Tian et al. [23] studied the elastic
anisotropy and ductility of TiZrVNb and TiZrNbMoVx (x = 01.5)
adopting the bcc structures using ab initio method.
In this work, we carry out ab initio SQS and CPA calculations to
establish the elastic properties of three refractory HEAs: TiZrVNb,
TiZrNbMo and TiZrVNbMo all crystalizing in the bcc structure. We
place special attention on the impact of LLD on the bulk properties and parallel establish the accuracy of single-site CPA for these
HEAs. The theoretical data is used to assess the rule of mixtures for
the lattice parameter and elastic constants. Alloying effects are discussed starting from the two base alloys (TiZrVNb and TiZrNbMo)
and focusing on the effect of Mo and V (TiZrVNbMo).
The rest of the paper is arranged as follows. The calculational
methods and related numerical details are given in Section 2 and
the two sets of ab initio data are compared in Section 3. The results
for the lattice parameter, mixing energy and elastic constants are
given in Section 4. Section 5 presents the discussions on the alloying

effect and rule of mixtures. The paper ends with a brief summary in
Section 6.
2. Computational method
2.1. Total-energy calculations
All calculations were performed within the framework of density functional theory formulated within the generalized gradient
approximation (GGA) by Perdew, Burke, Ernzerhof (PBE) [51] for
the exchange-correlation functional. The Kohn-Sham equations were
solved using the Vienna Ab initio Simulation Package (VASP) [5254]
and the Exact Mun-Tin Orbital (EMTO) methods [5558]. The substitutional disorder was treated using the SQS technique in VASP
calculations and the CPA in EMTO calculations.
In EMTO calculations, the single-electron equations were solved
with the scalar-relativistic approximations and soft-core scheme.
The EMTO total energy was obtained by the full charge density technique (FCD) [59,60], and the Kohn-Sham equations were solved for
the optimized overlapping mun-tin (OOMT) potential. The SQS
structures for VASP calculations were constructed using the Alloy
Theoretic Automated Toolkit (ATAT) [61,62]. The disordered SQS
supercells were either used as rigid bcc supercells (denoted by SQSu )
or taking into account the local chemical environment induced LLD
(SQSr ). Comparing the so obtained sets of data can give direct evidence for the impact of LLD on the computed physical properties. For
the VASP calculations, we adopted the projector augmented wave
(PAW) method [63].
2.2. Elastic constants
For bcc structures, there are three independent elastic constant:
C11 , C12 and C44 and the associated parameters of the tetragonal
shear modulus C = (C11 C12 )/2 and the bulk modulus B = (C11 +
2C12 )/3. In this work, all elastic constants were calculated according to the energy-strain relation. The theoretical equilibrium volume
and bulk modulus were derived from an exponential Morse-type
function [64] tted to the total energies calculated for nine different atomic volumes. In order to obtain the two cubic shear moduli C
and C44 , we used volume-conserving orthorhombic and monoclinic
deformations as described

0
0
1 + eo
0
,
1 eo
0
0
0
1/(1 eo2 )

(1)

and

0
1 em
.
em 1
0


2
0 0 1 / 1 em

(2)

The strains eo and em vary between 0 and 0.025 (0.05) for six
energy points for VASP (EMTO) calculations. In studies based on SQS,
the crystal symmetry is lowered due to the quasi-random distribution of the atomic species. Therefore, the elastic constants tensor has
a lower symmetry than cubic. Here we dropped the non-cubic components and made an arithmetical average between C11 , C12 and C44
calculated for the three main crystallographic directions following
the scheme described in Refs. [6567].
The polycrystalline shear modulus (G) was obtained according to
the Hill averaging method G = (GV + GR )/2, where the Reuss and
Voigt bounds are GR = 5(C11 C12 )C44 (4C44 + 3C11 3C12 ) 1 and
GV = (C11 C12 + 3C44 )/5. The Young modulus (E) and the Poisson
ratio (m) are connected to B and G by the relations E = 9BG/(3B + G)

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

and m = (3B 2G)/(6B + 2G). The Zener anisotropy ratio is dened


as AZ = 2C44 /(C11 C12 ).
2.3. Numerical details
The EMTO Greens function was calculated for 16 complex energy
points distributed exponentially on a semicircular contour including
the valence states. In the basis set s, p, d and f orbitals (lmax = 4) were
included. The number of orbitals for the full-charge density (FCD)
was chose as lhmax = 8. The electrostatic correction to the single-site
CPA was described using the screened impurity model [68] with
a screening parameter of 0.6. To obtain the accuracy needed for
the calculation of elastic constants we used about 19,68326,973
inequivalent k-points in the irreducible wedge of the orthorhombic
and monoclinic Brillouin zones.
The disordered SQS structures were generated with 128 atoms
(4 4 4) for TiZrVNb and TiZrNbMo and 250 atoms (5 5 5)
for TiZrVNbMo. The atomic positions of the SQS supercells can be
found in Ref. [69]. The VASP calculations were performed using an
energy cutoff of 400 eV for the plane wave basis set. Integrations
in the Brilouin zone were performed using k points generated with
5 5 5 mesh grids for the 128-SQS structures and 3 3 3 for
the 250-SQS structure. The changes of the total energies with the
number of k points and the cutoff energy were tested to ensure
10 2 eV/ convergence for the Hellmann-Feynman forces used for
the optimization of the atomic positions. The energy tolerance for
the electronic relaxation was set as 1 10 6 eV. For both rigid and
relaxed SQS structures, the volumes were optimized while keeping
the shape xed.
3. Assessing the accuracy
In order to assess the accuracy of our theoretical methods, rst we
test the lattice constants and elastic parameters for the ve elemental metals. The low-temperature crystal structure is hcp for Ti and
Zr and bcc for V, Nb, Mo and the three HEAs considered here [3,19].
With increasing temperature, both Ti and Zr undergo an hcp-bcc
structural transition. The experimental lattice parameters (for all
present metals) and bulk moduli (for V, Nb and Mo) quoted below are
those measured for the bcc phase, extrapolated to 0 K and corrected
for the zero point phonon effect. For details about the extrapolation
see Ref. [70] and references therein.
In Fig. 1, the VASP and EMTO lattice parameters and bulk moduli
of Ti, Zr, V, Nb and Mo are compared with the experimental values.
The data are also listed in Table 1. The two sets of theoretical lattice
constants are in excellent agreement with each other and also with
the experimental values. The EMTO bulk moduli are slightly smaller
than the VASP (Ti, Zr, V, Nb and Mo) and experimental values (Nb and
Mo), the largest relative deviations being around 11%. The lattice constants and bulk moduli follow the trends aV < aMo < aTi < aNb < aZr
and BMo > BV > BNb > BTi > BZr , respectively. That is, the bulk
moduli of the 3d and 4d metals, when considered separately, correlate well with the size of the corresponding lattice constants.
Fig. 2 compares the present theoretical elastic constants of elemental metals to the room-temperature experimental data. No
experimental values are given for bcc Ti and Zr (being unstable at
room temperature). The three sets of C11 values are in good agreement with each other, except for V where both methods slightly
underestimate the elastic constant. The agreement is less perfect for
the other two elastic constants. In particular, EMTO gives lower C12
for Nb and Mo relative to the VASP and experimental values. The
underestimated C12 by EMTO is reected in positive C = (C11
C12 )/2 for Ti (5 GPa) and Zr (7 GPa), which is in contrast to the
small but negative C values predicted by VASP for these two metals (8 and 3 GPa, respectively). We recall that mechanical stability
requires C11  C12 , meaning that EMTO predicts elastically barely

stable Ti and Zr. Nevertheless, the above absolute deviations between


VASP and EMTO C values are below those found for V, Nb and Mo.
The EMTO C44 values are higher than those by VASP and eventually
closer to the experimental values. For V, Nb and Mo, VASP underestimates C44 relative to the experimental values by 46%, 39% and 23%,
respectively, compared to 22%, 12% and 6% errors found for EMTO.
Based on the above ndings, we conclude that in general the two
ab initio methods provide consistent results for elemental metals.
Somewhat larger deviations between VASP and EMTO predictions
are observed in the case of C12 and C44 elastic constants. Since none
of the methods performs clearly better for all parameters when
compared to the experimental data, we continue to present results
obtained by both approaches and discuss the alloying effects and rule
of mixtures separately for the two sets of theoretical data.
4. Results
4.1. Lattice constant
Fig. 3 displays the lattice constants and bulk moduli of bcc
TiZrVNb, TiZrNbMo and TiZrVNbMo alloys calculated using the SQS
and CPA approximations. Numerical data are also listed in Table 2.
It is found that the CPA lattice constants are in perfect agreement
with the SQSu values obtained without relaxing the atomic positions.
The differences between these two sets of theoretical predictions
are 0.12%, 0.15% and 0.12%, respectively. Similar deviations are also
observed for the pure elements (see Table 1) and thus they may
be considered as the characteristic differences between EMTO and
VASP results for the lattice parameter. Actually, the same conclusion
holds also for the bulk modulus. CPA systematically underestimates
the SQSu bulk moduli of bcc TiZrVNb, TiZrNbMo and TiZrVNbMo (by
about 911 GPa), which is consistent with (even smaller than) the
deviations seen for pure metals (Fig. 1 (b)).
Comparing the results from the two sets of SQS calculations, we
nd that the lattice constants are increased by LLD for all three
HEAs. The differences between CPA and SQSr lattice parameters
are 0.40%, 0.48% and 0.43% for TiZrVNb, TiZrNbMo and TiZrVNbMo,
respectively. These deviations should still be considered very small
indicating that the present theoretical equilibrium lattice parameters are robust and they represent the correct DFT values obtained
at PBE level. LLD slightly increases the bulk moduli of all three HEAs.
Nevertheless, the changes are very small (below 3 GPa) so that the
good agreement between the three sets of theoretical bulk moduli
remains.
We conclude that the two ab initio approaches, CPA and SQS (with
and without LLD), predict very similar equation of states for the
present HEAs. Furthermore, LLD is found to have insignicant inuence on the equilibrium volume and bulk modulus of these HEAs,
despite of the nearly 20% difference between the largest and the
smallest lattice parameters of the alloy constituents.
4.2. Mixing energy
The enthalpy of formation of HEAs is one of the key parameters
controlling the solid solution phase formation [26]. For the present
alloys, the estimated enthalpy of formation are 0.03 (TiZrVNb),
2.50 (TiZrNbMo) and 2.72 (TiZrVNbMo) kJ/mol [73]. These values
are all located inside the empirical limits from 15 to +5 kJ/mol [74]
for solid solution formation. Hence, it is expected that these systems
form a solid solution.
The enthalpy of formation is dened with respect to the ground
state structure of the alloy constituents, i.e. hcp for Ti and Zr, and
bcc for V, Nb and Mo. Here we focus on the mixing energy dened
with respect to the total energies of the alloy components in the
bcc structure. In this way, when comparing different theoretical
results we can exclude effects coming from the hcp-bcc structural

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
4

L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

(b)

(a)

Fig. 1. Comparison between the theoretical (VASP and EMTO) and the experimental (extrapolated to static conditions [70]) lattice parameters (in ) (a) and bulk moduli (in GPa)
(b) of bcc Ti, Zr, V, Nb and Mo.

We observe that the fully relaxed mixing energies follow the


trend DE (TiZrNbMo) < DE(TiZrVNbMo) < DE(TiZrVNb), which
is different from the above quoted trend of the formation
enthalpies [73]. The present VASP (EMTO) total energy differences
between bcc and hcp Ti and Zr are 10.4 (8.1) and 7.7 (5.5) kJ/mol,
respectively. These gures are in line with the Local Density Approximation results reported previously [75]. Using the above VASP
bcc-hcp structural energy differences in combination with the SQSu
mixing energies, for the formation enthalpies of TiZrVNb, TiZrNbMo
and TiZrVNbMo we get 11.1, 0.15 and 6.3 kJ/mol, respectively. The
rather high formation enthalpy of TiZrVNb may indicate that the solid
solution formation is at least questionable for this HEA in the light of
the empirical upper limit of 5 kJ/mol by Yang et al. [74]. We notice
that the CPA and SQSu mixing energies yield formation enthalpies far
above the Yangs upper limit. Hence, it is clear that without proper
account for the LLD effects no meaningful theoretical prediction can
be made on the thermodynamics of HEAs. Finally, we would like to
note that the above-quoted empirical solid solution formation limits
have to be treated with precautions in the mirror of recent developments by Laurent-Brocq et al. [76]. Probing the present systems
against phase separation, segregation or intermetallic formation is,
however, beyond the scope of this work.

energy differences between various methods. We dene the mixing


energy as
DE = Ealloy

ci Eibcc

(3)

where Ealloy is the total energy per atom of the HEA and Eibcc is the
equilibrium energy of the ith alloy component calculated for the bcc
structure. ci stands for the concentration.
The calculated mixing energies for TiZrVNb, TiZrNbMo and
TiZrVNbMo alloys are shown in Fig. 4. One can see that DE calculated from CPA and SQSu are in reasonable agreement with each
other. The largest deviation of 4 kJ/mol (corresponding approximatively to 25% difference) is obtained for TiZrVNb. The differences
between SQSu and CPA data are due to the single-site approximation
in CPA. Here we use the screened impurity model [68] with a xed
screening parameter to describe the electrostatic potential within
CPA. Adjusting the screening parameter, one could in principle reproduce with very high accuracy the SQSu results by CPA. On the other
hand, the DE values obtained by the rigid-lattice SQSu approximation turn out to be far too large when compared to the relaxed SQSr
results, which makes such tuning of the screening parameter rather
useless. As seen in Fig. 4, the theoretical DE values are signicantly
decreased when we take into account the LLD. Actually, the LLD
effect on the mixing energies is substantially larger than the CPASQSu differences, which makes the CPA results for the mixing energy
less reliable.

4.3. Elastic parameters


Fig. 5 displays the single-crystal elastic constants (Cij ) from CPA
and both sets of SQS calculations. The CPA and the average SQSu
and SQSr values are listed in Table 2 as well. The individual SQS

Table 1
Theoretical (EMTO and VASP) equilibrium lattice parameters a () and elastic constants (GPa) for Ti, Zr, V, Nb and Mo in bcc structure. Shown are also the high-temperature (Ti
and Zr) and room-temperature (V, Nb and Mo) experimental lattice parameters and bulk moduli extrapolated to 0 K and corrected for zero point phonon effect (see Ref. [70] and
references therein). The experimental single-crystal elastic constants are from Ref. [71].

Ti

Zr

Nb

Mo

Method

C11

C12

C44

EMTO
VASP
Expt.
EMTO
VASP
Expt.
EMTO
VASP
Expt.
EMTO
VASP
Expt.
EMTO
VASP
Expt.

3.258
3.253
3.265
3.574
3.574
3.571
2.998
3.001
3.024
3.310
3.310
3.294
3.165
3.169
3.141

104.5
107.9

82.3
91.0

174.9
184.0
161
155.5
171.4
174
247.2
264.4
278

111.2
96.8

92.3
87.0

279.6
266.2
232
252.1
243.7
253
476.0
454.2
450

101.2
113.5

77.3
92.9

122.5
142.9
119
107.1
135.3
133
132.8
169.2
173

64.9
43.6

49.2
28.2

36.1
24.9
46
34.6
19.0
31
117.3
96.0
125

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

(a)

(a)

(b)

(b)

Fig. 3. Theoretical lattice constants of the underlying bcc unit cell (, upper panel)
and bulk moduli (GPa, lower panel) obtained for TiZrVNb, TiZrNbMo and TiZrVNbMo
alloys using the EMTO-CPA and VASP-SQS methods. For the latter, both the unrelaxed
(SQSu ) and relaxed (SQSr ) results are shown. The experimental data for TiZrVNb and
TiZrVNbMo are taken from Ref. [48] and Ref. [72], respectively.

(c)

Fig. 2. Comparison between theoretical (VASP and EMTO) and experimental (room
temperature) C11 (a), C12 (b) and C44 (c) of bcc Ti, Zr, V, Nb and Mo.

elastic constants due to the low symmetry induced by the chemical quasi-randomness in SQS calculations are provided in Table 3.
By comparing the individual values one can learn about the effect of
reduced-symmetry on the elastic parameters. For instance, the deviation is about 12.8 GPa for C44 of TiZrNbMo, representing about 65%
of the average value. Sizable scatter is seen for C11 as well, especially for TiZrVNb. One expects that increasing the size of the SQS
cell should reduce the differences between the individual elastic constants and eventually restore the cubic symmetry for very large cells.
In the following we discuss only the average SQS elastic constants.
According to the numerical values in Table 2, we nd that the
mechanical stability criteria (C > 0, C44 > 0, B > 0) are fullled by all three sets of data and thus the TiZrVNb, TiZrNbMo and
TiZrVNbMo alloys are predicted to be dynamically stable. However,
when comparing the CPA and SQS results, we nd that some elastic
constants are sensitive to the employed method. For example, C and
C44 obtained in CPA calculations are by 15 and 30 GPa, respectively, larger than those calculated using the SQS method. We should
recall that similar differences between VASP and EMTO single-crystal
elastic constants are found for the pure metals as well (Table 1). This

observation suggests that the main deviations between the CPA and
SQS values in Table 2 in fact originate from the errors associated
with the underlying DFT methods rather than from the way how the
chemical randomness is treated by CPA or SQS. This is in line with
previously reported conclusions. For instance, Zaddach et al. studied
the elastic constants of CoCrFeMnNi 3d-high entropy alloy based on
the EMTO-CPA and VASP-SQS methods [77]. They found the elastic
coecients in EMTO-CPA calculations in close agreement to those
obtained in the VASP calculations for ferromagnetic conguration.
Niu et al. compared the elastic constants of non-equiatomic NiFeCrCo
high entropy alloy (i.e., 1040 at.%) by the EMTO and VASP [35]. Good
agreement was found between the two methods.
The lattice distortion induced by the size of different elements
can affect the strength of bcc HEAs [78]. Therefore, it is essential to
consider the inuence of LLD on the mechanical properties of HEAs.
Our results can reveal this effect in the case of elastic parameters.
Comparing the SQSu and SQSr data in Table 2, it is concluded that the
elastic constants of the present HEAs remain almost unchanged upon
taking into account the LLD. For instance, the average C11 increases
by 2.9 GPa (1.8%), 4.5 GPa (2.3%) and 5.6 GPa (2.7%) for TiZrVNb,
TiZrNbMo and TiZrVNbMo, respectively, with local lattice relaxation.
The relative LLD induced changes in C44 are somewhat larger due the
small absolute value.
In Table 2, the Zener anisotropy ratio AZ = C44 /C , the Young
modulus E, the shear modulus G, the Cauchy pressure (C12 C44 ), the
Pugh ratio B/G and the Poisson ratio m are also listed. Unfortunately,
no experimental information on these elastic moduli is available. In
general the SQS and CPA values for elastic moduli are consistent with
each other. The large differences obtained for the shear and Young
moduli are primarily due to the strong underestimation of the C44
elastic constant of pure metals (also HEAs) by VASP as compared to
the EMTO and experimental data (Table 1). Notice that all HEAs considered here are relatively isotropic (AZ is close to 1), which places G
close to both single-crystal shear elastic constants (C and C44 ).
The B/G ratio is often used as an indicator of the brittle-ductile
behavior. High values of B/G (1.75) indicate that the material is
ductile, while low values are associated with brittleness. The Poisson ratio m may also be used to predict the brittle-ductile behavior. It

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
6

L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

Table 2
The theoretical lattice constant for the bcc underlying lattice (a, ), elastic constants C11 , C12 , C44 and C (GPa); the Zener anisotropy (AZ = C44 /C ) and Cauchy pressure (C12 C44 )
(GPa), polycrystalline elastic moduli (B, G, E, GPa), Poissons ratio (m) and the Pugh ratio (B/G) of the TiZrVNb, TiZrNbMo and TiZrVNbMo alloys calculated using the EMTO-CPA and
VASP-SQS methods. For SQS both unrelaxed (SQSu ) and relaxed (SQSr ) results are shown. For reference, we also list the CPA results (CPAa ) obtained by Tian et al. [23].
Method
TiZrVNb
CPA
CPAa
SQSu
SQSr
TiZrNbMo
CPA
CPAa
SQSu
SQSr
TiZrVNbMo
CPA
CPAa
SQSu
SQSr

C11

C12

C44

C

AZ

C12 C44

B/G

3.290
3.281
3.294
3.303

165.8
166.4
156.9
159.8

92.8
94.7
111.7
114.3

50.5
53.8
21.6
18.5

36.5
35.9
22.6
22.8

1.385
1.500
0.956
0.810

42.2
41.0
90.0
95.8

117.1
118.6
126.7
129.5

44.4
45.7
22.0
20.1

118.2
121.1
62.4
57.3

0.332
0.330
0.418
0.426

2.639
2.604
5.756
6.447

3.306
3.304
3.311
3.322

209.6
209.9
198.9
203.4

98.8
101.0
122.1
121.5

49.9
52.6
26.9
29.6

55.4
54.4
38.4
41.0

0.901
0.966
0.702
0.723

48.9
48.4
95.1
91.8

135.8
137.3
146.9
148.2

52.0
53.3
31.1
33.8

138.4
141.7
87.1
94.1

0.330
0.328
0.401
0.394

2.608
2.575
4.729
4.408

3.252
3.248
3.256
3.266

216.3
213.7
203.7
209.3

100.2
100.7
121.1
123.0

47.8
50.9
24.1
26.9

58.1
56.5
41.3
43.1

0.824
0.900
0.583
0.623

52.4
49.8
97.1
96.2

138.9
138.5
148.7
151.8

51.7
53.2
29.9
32.5

137.9
141.1
84.1
91.0

0.335
0.330
0.406
0.400

2.688
2.608
4.969
4.669

has been reported that bulk metallic glasses with m > 0.31 are ductile. Similarly, positive Cauchy pressure indicates ductility as well.
According to Table 2, for the present systems B/G is always larger
than the critical value of 1.75, m is above 0.31 and the Cauchy pressure is positive. Therefore, the TiZrVNb, TiZrNbMo and TiZrVNbMo
HEAs are predicted to be ductile by both CPA and SQS calculations.
The above phenomenological ductility indicators are very weakly
affected by LLD.
4.4. Local lattice distortions
The magnitudes of the LLDs are directly evaluated from the
atomic coordinates obtained through the VASP-SQS geometry optimization. For each relaxed HEA system, a histogram of the radial
distributions for every atom in the cell is presented in Fig. 6. Signicant distortion may be observed in the relaxed cells, which results in
an overlap of the rst and second nearest neighbor (NN) shells represented in the histograms. This overlap is more likely in the present
alloys due
to the close proximity of the rst two NN shells in the bcc
lattice ( 3ao /2 and ao ). Before the LLDs are analyzed, we will briey
describe the procedure by which the results were obtained.
The histograms in Fig. 6 were generated by calculating the distance to each atom around every site and retaining only those within
a radial cut off of approximately 4 . The histograms count atoms in
bins of width 0.03 between 2.49 and 3.69 . Because this prescription double counts every pair, the number of atoms in each bin was
then reduced by half. Due to the overlap of NN shells in the relaxed

Fig. 4. Theoretical mixing energies of bcc TiZrVNb, TiZrNbMo and TiZrVNbMo alloys
obtained by unrelaxed SQS (SQSu ), relaxed SQS (SQSr ) and CPA calculations. In all cases
the reference states are the pure elements in bcc structure.

systems, the distances determined from the relaxed supercells are by


themselves insucient to determine whether a given pair belongs
in the rst or second NN shell. Here, we use the ideal unrelaxed SQS
to determine the indices of rst and second NNs, i.e. we make use
of the unrelaxed indexing to associate the atoms to initial coordination shells after lattice relaxation. These indices are then used to
identify the pair distances plotted in Fig. 6. Pairs of atoms that were
rst NNs in the unrelaxed system are binned in one histogram, while
those that were second NNs in the unrelaxed system are binned in a
separate histogram, both of which are represented on the same plot.
It should be noted that the standard deviations determined for the
radial distribution of all sites for the rst and second NNs in each system are identical to 4 decimal places, which should be the case for
a periodic bulk supercell and indicates the degree of relaxation from
the ideal lattice.
From inspection of Fig. 6, it is clear that the most signicant LLDs
are found in the TiZrVNb HEA. Smaller, yet still signicant, LLDs are
found in TiZrNbMo and TiZrVNbMo alloys. It is evident in Fig. 6 that
the radial distribution histograms for rst and second NN pairs in
TiZrVNb are more spread out than the corresponding histograms for
the other HEAs. Moreover, the histograms for TiZrVNb seem to be
superpositions of small subpeaks, whereas the histograms for the

Fig. 5. Single-crystal elastic constants of bcc TiZrVNb, TiZrNbMo and TiZrVNbMo


alloys obtained by unrelaxed SQS (SQSu ), relaxed SQS (SQSr ) and CPA calculations. For
SQS, shown are the average values.

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

Table 3
The relevant elastic parameters of TiZrNbMo, TiZrVNb and TiZrVNbMo alloys calculated using VASP-SQS method. The unrelaxed data are listed in the parentheses.
Alloy

TiZrVNb
TiZrNbMo
TiZrVNbMo

C11

C12

C44

C11

C22

C33

C12

C13

C23

C44

C55

C66

169.3
(160.0)
199.6
(197.6)
213.4
(205.0)

153.9
(154.9)
208.5
(201.4)
213.2
(203.7)

153.4
(155.9)
202.1
(197.7)
201.3
(202.4)

109.5
(110.1)
123.3
(121.6)
121.0
(121.3)

117.5
(112.2)
122.1
(121.6)
127.0
(121.4)

117.2
(112.7)
118.9
(123.0)
121.1
(120.7)

18.5
(21.0)
32.8
(32.6)
20.2
(21.2)

18.8
(20.8)
34.3
(28.5)
32.0
(25.2)

18.1
(23.0)
21.8
(19.8)
26.9
(25.7)

other HEAs are more centered around a single peak. In order to


understand the more signicant distortion found in TiZrVNb as compared to TiZrNbMo and TiZrVNbMo, we have calculated the average
pair distance and standard deviation for rst NN pairs at every site
and also at each site of a given element in each of these alloys. These
results are presented in Table 4. A similar approach based on the
matrix description of the LLDs effect in bcc MoNbTaVW high entropy
alloy was recently used in Ref. [36].
Each four-component alloy has 32 unique chemical sites, whereas
there are 50 unique chemical sites in the ve-component alloy. With
8 rst NNs in the bcc unit cell, this leads to sample sizes of 256 and
400 for the four- and ve-component systems, respectively. Looking
at the results in Table 4, it is apparent that within each HEA there is
little variation of the standard deviations for different chemical sites.
The average radial distance around each site, however, varies more
signicantly from site to site. Generally, the average radial distance

is largest around the largest atom, Zr, and smaller around the smaller
atoms Mo and V. Ti and Nb are intermediate in average radial distance as compared to the other elements, and are generally closer in
average radial distance to the overall average of the bulk.
The more signicant standard deviations in TiZrVNb originates
from the variation in average radial distances around each atomic
site. This is a product of both atomic size and the resultant electronic
structure of the alloy. TiZrVNb and TiZrNbMo are very close in lattice
spacing, which is clear from inspection of the ideal rst and second
NN distances in Fig. 6 or the average radial distance values of all
sites in Table 4. The average radial distance for rst NN pairs around
V sites is smaller as compared to Mo sites. Comparison of the fourcomponent alloys shows that Ti and Nb have smaller average radial
distances in the alloy containing V, while Zr has a slightly larger average radial distance in the alloy containing V. This is consistent with
Ref. [36] and it is due to the smaller V atom relative to the rest of the
alloy constituents.
If we envision the histograms in Fig. 6 as the superposition of
a set of element-specic histograms for radial distributions, we can
readily explain their shapes. In such a picture, an average value in
Table 4 would represent the location of the peak of the distribution around a given element, while the standard deviation would
represent the width of that peak. We see from Table 4 that the average values, and therefore the peak locations, are more spread out
for TiZrVNb than for the other two HEAs; and the standard deviations, that is, the width of each peak, are higher. Therefore, we expect
the overall distribution of radial pair distances in TiZrVNb to feature multiple broad peaks, whereas the overall distributions for the
other two HEAs should feature relatively narrow peaks at roughly
the same location. This is exactly what we observe in Fig. 6: multiple
at peaks for the rst NN distribution in TiZrVNb, and a single narrow peak for the rst NN distributions in TiZrNbMo and TiZrVNbMo.

Table 4
Average distances (ravg ) and the standard deviations (s) for the rst nearest neighbor
(NN) pairs around each chemical site in TiZrVNb, TiZrNbMo, and TiZrVNbMo HEAs.
System

Site

1st NN ravg s

TiZrVNb

All
Ti
Zr
V
Nb
All
Ti
Zr
Nb
Mo
All
Ti
Zr
V
Nb
Mo

2.866 0.123
2.875 0.111
2.946 0.111
2.785 0.107
2.856 0.106
2.880 0.090
2.884 0.072
2.944 0.080
2.872 0.082
2.819 0.072
2.832 0.098
2.850 0.085
2.914 0.088
2.782 0.087
2.833 0.077
2.780 0.087

TiZrNbMo

TiZrVNbMo
Fig. 6. Histogram plot of LLDs for TiZrVNb (top), TiZrNbMo (middle), and TiZrVNbMo
(bottom). The 1st and 2nd nearest neighbors (NNs) are colored red and gray, respectively. The ideal (unrelaxed) positions of the rst and second NNs based on the lattice
parameter are indicated with dashed lines. (For interpretation of the references to
color in this gure legend, the reader is referred to the web version of this article.)

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
8

L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

Table 5
Changes in the single-crystal elastic constants (GPa) as going from TiZrVNb or
TiZrNbMo to TiZrVNbMo. TiZrVNbMo (TiZrNbMoV ) stands for the effect of Mo (V) on
the elastic constants of TiZrVNb (TiZrNbMo). Results are shown for CPA, SQSu and SQSr
(all data taken from Table 2).

CPA
TiZrVNbMo
TiZrNbMoV
SQSu
TiZrVNbMo
TiZrNbMoV
SQSr
TiZrVNbMo
TiZrNbMoV

DC11

DC12

DC44

DC

D(C12 C44 )

50.5
6.7

7.4
1.4

2.7
2.1

21.5
2.7

10.2
3.5

46.8
4.8

9.4
1.0

2.5
2.8

18.7
2.9

7.1
2.0

49.5
5.9

8.7
1.5

8.4
2.7

20.3
2.1

0.4
4.4

A similar analysis applies to the shape observed for the histograms


of the second NN distributions.
In summary, signicant LLDs are found in TiZrVNb, TiZrNbMo,
and TiZrVNbMo. These have been illustrated by the presentation of
radial distribution histograms as well as quantitative analysis. How
the LLDs inuence other physical properties of the alloys will be
analyzed through the use of both the EMTO-CPA and the VASP-SQS
approaches.
5. Discussion
5.1. Alloying effect
Table 5 lists the calculated changes in the single-crystal elastic
constants for TiZrVNbMo relative to TiZrVNb and TiZrNbMo. These
changes reveal the HEA-type of alloying effect when going from a
four component to a ve component equiatomic alloy. We should
emphasize that this process means that the atomic fraction of all constituents in the base alloy decreases from 0.25 to 0.20 when 20% V
or Mo is introduced. In Table 5 alloying effects are shown for CPA,
SQSu and SQSr . It is obvious that adding equi-concentration V (Mo)
to TiZrVNb (TiZrNbMo) strongly affects some of the elastic constants.
In all three sets of theoretical data, the largest changes are
obtained for C11 when Mo is added to TiZrVNb. Actually, the theoretical C11 of Mo is much larger than that of TiZrVNb, which could partly
explain the large positive slope. The nearly 50 GPa increase in C11
upon Mo addition to TiZrVNb is reected by DC being about half of
DC11 . The second largest change is seen for C12 again upon Mo addition, which is also in line with the individual elastic constants of Mo
and TiZrVNb. The changes are much smaller for C44 and for all elastic constants when V is added to TiZrNbMo. The alloying effects by V
may be understood on the same footing as above, i.e. by comparing
the elastic parameters of pure V to those of the TiZrNbMo host.
The trend of C44 is much less obvious when Mo is introduced in
TiZrVNb. Pure Mo has the largest C44 among all metals considered
here (96.0 GPa in VASP and 117.3 GPa in EMTO, Table 1) whereas
C44 of TiZrVNb (18.521.6 GPa in SQS and 50.5 GPa in CPA, Table 2)
is similar to those of the other two HEAs. Simple linear mixing for
C44 of TiZrVNbMo predicts 35.036.5 GPa in VASP-SQS and 63.7 GPa
in EMTO-CPA. If one uses C44 of Mo computed at the equilibrium
volume of TiZrVNb (72.4 GPa), linear mixing still yields large positive change for C44 when going from TiZrVNb to TiZrVNbMo. On this
ground, one might expect that adding equiatomic Mo to the base
alloy should strongly enhance C44 . However, ab initio calculations
show a completely different trend. Actually, both CPA and SQSu give
small changes (2.7 and 2.5 GPa, respectively) and the somewhat
larger DC44 predicted by SQSr (8.4 GPa) is also only 50% of what
the linear mixing suggests. The reason behind this inconsistency
should be associated with the complex chemical interactions within

the HEAs which are to some extend captured by both alloy theories (CPA and SQS) but fall outside of the linear rule of mixtures. In
the next section, we make a more robust assessment of the rule of
mixtures for the present HEAs and investigate if at least the general
trends could be reproduced by such simple estimate.
5.2. Rule of mixtures for refractory HEAs
The rule of mixtures have often been employed to make an initial screening of the properties of alloys and compounds. It gives a
simple estimate of the selected and often unknown property of the
alloy based on the accessible properties of the alloy constituents. A
widely used example is Vegards rule where the lattice constant of a
binary alloy is estimated form a linear interpolation between the lattice parameters of the constituents (assuming same crystal lattice for
both). Here we possess sucient amount of reliable ab initio data for
the alloys and their components to be able to test such rules in the
case of refractory HEAs.
Based on data obtained for the elemental metals, we estimated
the lattice constants and elastic parameters of HEAs according to

pest =

N
1
pi ,
N

(4)

i=1

where pi stands for elastic parameters or lattice parameter of pure


metal i in bcc structure and N is the number of components in the
alloy. We compare the so estimated physical parameter to the one
computed in fully self-consistent ab initio calculation. The results are
shown in Fig. 7. Upper panel contains data obtained using the EMTO
method for pure metals and the EMTO-CPA method for HEAs. Lower
panel shows results obtained by VASP and VASP-SQS both without
(SQSu ) and with (SQSr ) LLD. Since bcc Zr and bcc Ti are dynamically
unstable (at static conditions) one cannot dene a physically meaningful shear modulus for these two hypothetical metals. Hence, G
(same applies to E) cannot be obtained directly from the rule of mixture. But these polycrystalline elastic parameters can be computed
from the averaged single-crystal data. The so estimated and calculated values are also shown in Fig. 7. Note that the Poisson ratio can
easily be derived from G and E.
We observe that in general the trends of the lattice parameters
and elastic constants are well reproduced by the rule of mixtures. The
largest errors (deviations between calculated and estimated data) are
obtained for the lattice parameters. The differences are not systematic: Eq. (4) overestimates the CPA lattice parameters of TiZrNbMo
and TiZrVNbMo and slightly underestimates that of TiZrVNb. Very
similar deviations are found for the SQS lattice parameters as well.
On the other hand, the estimated elastic constants are surprisingly
close to the theoretical values in both CPA and SQS calculations. One
can conclude that the linear rule of mixtures provides a rather useful estimate for the elastic parameters of unknown HEAs assuming
that the elastic constants of the alloy constituents are known for
the same crystal structure. Nevertheless, one should notice that this
nice agreement strictly holds only on the scale of the present elastic constants (0250 GPa), but fails when one looks at the individual
elastic parameters on their own scales.
6. Conclusions
The elastic properties of bcc TiZrVNb, TiZrNbMo and TiZrVNbMo
random alloys were studied using two rst-principles methods in
combination with two techniques for random alloys. First, the DFT
tools (VASP and EMTO) were assessed by comparing the two sets
of ab initio bulk parameters with the available experimental data
for pure elements. The two methods provide consistent equation of
states for bcc Ti, Zr, V, Nb, and Mo, but systematic deviations are

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

(a)

(b)

The magnitude of the LLDs were measured for the three HEA systems here using histograms for showing the data. Here we ultimately
were interested in quantifying the degree of the distortion. An overlap of the rst and second nearest neighbor shell was observed
in TiZrVNb, TiZrNbMo and TiZrVNbMo. That means that signicant
distortions exist in the relaxed cells.
We also studied the alloying effects and found large deviations from the trends expected from simple estimates. In particular,
when equiatomic Mo is added to TiZrVNb, the C44 elastic parameter decreases or slightly increases (depending on the employed DFT
solver and alloy theory) despite the fact that Mo has the largest C44
among all metals and alloys considered here. Having computed all
necessary data for the pure end members and alloys, we had the
possibility to assess the rule of mixtures for HEAs. The estimated
elastic parameters are found to be rather close to the values obtained
from the CPA and SQS calculations especially on the scale of the
present values (0250 GPa). But this simple estimate turned out to
be too rough to resolve delicate alloying effects like the one seen
for C44 of TiZrNbMo and TiZrVNbMo. Nevertheless, when based on
data obtained for the same crystal lattice, the rule of mixtures can
provide very useful rst-level estimates for the values of the lattice
parameters and elastic constants including general trends.

Acknowledgments
The authors acknowledge the Swedish Research Council (VR), the
Swedish Foundation for Strategic Research (SSF), the Swedish Foundation for International Cooperation in Research and Higher Education (STINT), the Carl Tryggers Foundations, the Swedish Innovation
Agency (VINNOVA), the Hungarian Scientic Research Fund (OTKA
109570), and the China Scholarship Council for nancial supports.
The computations were performed on resources provided by the
Swedish National Infrastructure for Computing (SNIC) at Linkping.

References

Fig. 7. Comparison between calculated and estimated single-crystal and polycrystalline elastic constants and lattice parameters of TiZrVNb, TiZrNbMo and TiZrVNbMo
HEAs. Upper panel shows data obtained with EMTO-CPA method and lower panel with
VASP-SQS both with and without LLD.

observed for the single-crystal elastic constants, especially for C12


and C44 . None of these methods in combination with the present DFT
approximation (PBE) is able to provide highly accurate results for all
elastic parameters. This means that alloying effects should always be
discussed with respect to the data for end members obtained using
the same underlying DFT tool.
In general, the ab initio results obtained with CPA and SQS for the
present HEAs are consistent with each other. Local lattice distortions
were investigated by use of SQS cells and the VASP method. LLD has
signicant impact on the mixing energy. Neglecting this effect can
lead to strongly overestimated enthalpy of formation and thus such
rough data is completely useless to draw any conclusions concerning
the solid solution formation. On the other hand, LLD has very small
effect on the equation of state and elastic constants. For these properties, CPA is clearly superior to SQS as it gives more exibility for
the number of components and deviations from equiatomic compositions. The two DFT tools in combination with the two alloy theories
eciently complement each other in ab initio description of complex
alloys.

[1] Y.Y. Chen, T. Duval, U.D. Hung, J.W. Yeh, H.C. Shih, Microstructure and electrochemical properties of high entropy alloys-a comparison with type-304
stainless steel, Corros. Sci. 47 (9) (2005) 22572279.
[2] H.P. Chou, Y.S. Chang, S.K. Chen, J.W. Yeh, Microstructure, thermophysical and
electrical properties in AlxCoCrFeNi high-entropy alloys, Mater. Sci. Eng. B 163
(3) (2009) 184189.
[3] V. Dolique, A.L. Thomann, P. Brault, Y. Tessier, P. Gillon, Thermal stability of
AlCoCrCuFeNi high entropy alloy thin lms studied by in-situ XRD analysis,
Surf. Coatings Technol. 204 (12-13) (2010) 19891992.
[4] O.N. Senkov, C.F. Woodward, Microstructure and properties of a refractory
NbCrMo0.5 Ta0.5 TiZr alloy, Mater. Sci. Eng. A 529 (2011) 311320.
[5] Y.J. Hsu, W.C. Chiang, J.K. Wu, Corrosion behavior of FeCoNiCrCux high-entropy alloys in 3.5% sodium chloride solution, Mater. Chem. Phys. 92 (1) (2005)
112117.
[6] J.M. Wu, S.J. Lin, J.W. Yeh, S.K. Chen, Y.S. Huang, H.C. Chen, Adhesive wear
behavior of AlxCoCrCuFeNi high-entropy alloys as a function of aluminum
content, Wear 261 (5-6) (2006) 513519.
[7] J.P. Couzini, G. Dirras, L. Perrire, T. Chauveau, E. Leroy, Y. Champion, I. Guillot,
Microstructure of a near-equimolar refractory high-entropy alloy, Mater. Lett.
126 (2014) 285287.
[8] O.N. Senkov, G.B. Wilks, D.B. Miracle, C.P. Chuang, P.K. Liaw, Refractory high-entropy alloys, Intermetallics 18 (9) (2010) 17581765.
[9] O.N. Senkov, J.M. Scott, S.V. Senkova, D.B. Miracle, C.F. Woodward, Microstructure and room temperature properties of a high-entropy TaNbHfZrTi alloy, J.
Alloys Compd. 509 (20) (2011) 60436048.
[10] Y.J. Zhou, Y. Zhang, Y.L. Wang, G.L. Chen, Microstructure and compressive properties of multicomponent Alx(TiVCrMnFeCoNiCu)100-x high-entropy alloys,
Mater. Sci. Eng. A 454 (2007) 260265.
[11] Y.J. Zhou, Y. Zhang, Y.L. Wang, G.L. Chen, Solid solution alloys of AlCoCrFeNiTix
with excellent room-temperature mechanical properties, Appl. Phys. Lett. 90
(18) (2007) 181904.
[12] Y. Zhou, Y. Zhang, F. Wang, Y. Wang, G. Chen, Effect of Cu addition on the
microstructure and mechanical properties of AlCoCrFeNiTi0.5 solid-solution
alloy, J. Alloys Compd. 466 (1) (2008) 201204.
[13] L.I. Anmin, X. Zhang, Thermodynamic analysis of the simple microstructure of
AlCrFeNiCu high-entropy alloy with multi-principal elements, Acta Metall. 22
(3) (2009) 219224.

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

ARTICLE IN PRESS
10

L-Y. Tian et al. / Materials and Design xxx (2016) xxxxxx

[14] B. Cantor, I.T.H. Chang, P. Knight, A.J.B. Vincent, Microstructural development in


equiatomic multicomponent alloys, Mater. Sci. Eng. A 375-377 (1-2 SPEC. ISS.)
(2004) 213218.
[15] M.S. Lucas, G.B. Wilks, L. Mauger, J.A. Munoz, O.N. Senkov, E. Michel, J. Horwath,
Absence of long-range chemical ordering in equimolar FeCoCrNi, Appl. Phys.
Lett. 100 (25) (2012) 251907.
[16] T.T. Shun, C.H. Hung, C.F. Lee, Formation of ordered/disordered nanoparticles in
FCC high entropy alloys, J. Alloys Compd. 493 (1) (2010) 105109.
[17] M.C. Gao, B. Zhang, S.M. Guo, J.W. Qiao, J.A. Hawk, High-entropy alloys in
hexagonal close-packed structure, Metall. Mater. Trans. A (2016) 111.
[18] J.W. Yeh, Recent progress in high-entropy alloys, Ann. Chim. Sci. Des Mater. 31
(6) (2006) 633648.
[19] J.W. Yeh, S.K. Chen, S.J. Lin, J.Y. Gan, T.S. Chin, T.T. Shun, C.H. Tsau, S.Y. Chang,
Nanostructured high-entropy alloys with multiple principal elements: novel
alloy design concepts and outcomes, Adv. Eng. Mater. 6 (2004) 299.
[20] Y. Zhang, T.T. Zuo, Z. Tang, M.C. Gao, K.A. Dahmen, P.K. Liaw, Z.P. Lu, Microstructures and properties of high-entropy alloys, Prog. Mater. Sci. 61 (October 2013)
(2014) 193.
[21] E. Fazakas, V. Zadorozhnyy, L.K. Varga, A. Inoue, D.V. Louzguine-Luzgin, F. Tian,
L. Vitos, Experimental and theoretical study of Ti20 Zr20 Hf20 Nb20 X20 (X = V or
Cr) refractory high-entropy alloys, Int. J. Refract. Met. Hard Mater. 47 (2014)
131138.
[22] Y. Zhang, X. Yang, P. Liaw, Alloy design and properties optimization of high-entropy alloys, JOM 64 (7) (2012) 830838.
[23] F. Tian, L.K. Varga, N. Chen, J. Shen, L. Vitos, Ab initio design of elastically
isotropic TiZrNbMoVx high-entropy alloys, J. Alloys Compd. 599 (2014) 1925.
[24] F. Tian, D. Wang, J. Shen, Y. Wang, An ab initio investgation of ideal tensile
and shear strength of TiVNbMo high-entropy alloy, Mater. Lett. 166 (2016)
271275.
[25] S. Huang, W. Li, X. Li, S. Schnecker, L. Bergqvist, E. Holmstrm, L.K. Varga,
L. Vitos, Mechanism of magnetic transition in FeCrCoNi-based high entropy
alloys, Mater. Des. 103 (2016) 7174.
[26] Y. Zhang, Y.J. Zhou, J.P. Lin, G.L. Chen, P.K. Liaw, Solid-solution phase formation
rules for multi-component alloys, Adv. Eng. Mater. 10 (6) (2008) 534538.
[27] S. Guo, C.T. Liu, Phase stability in high entropy alloys: formation of solid
solution phase or amorphous phase, Prog. Nat. Sci. Mater. Int. 21 (6) (2011)
433446.
[28] X. Yang, Y. Zhang, Prediction of high-entropy stabilized solid-solution in multi-component alloys, Mater. Chem. Phys. 132 (2) (2012) 233238.
[29] M.G. Poletti, L. Battezzati, Electronic and thermodynamic criteria for the occurrence of high entropy alloys in metallic systems, Acta Mater. 75 (2014)
297306.
[30] I. Toda-Caraballo, P.E.J. Rivera-Daz-del Castillo, A criterion for the formation of
high entropy alloys based on lattice distortion, Intermetallics 71 (2016) 7687.
[31] W. Voigt, Ueber die Beziehung zwischen den beiden Elasticitts constanten
isotroper Krper, Annalen der Physik 274 (12) (1889) 573587.
[32] A. Reuss, Calculation of the ow limits of mixed crystals on the basis of the
plasticity of monocrystals, Z. Angew. Math. Mech 9 (1929) 4958.
[33] R. Hill, The elastic behavior of a crystalline aggregate, Proc. Phys. Soc. Section A
65 (5) (1952) 349354.
[34] A.R. Denton, N.W. Ashcroft, Vegards law, Phys. Rev. A 43 (6) (1991) 31613164.
[35] C. Niu, A.J. Zaddach, C.C. Koch, D.L. Irving, First principles exploration of
near-equiatomic NiFeCrCo high entropy alloys, J. Alloys Compd. 672 (2016)
510520.
[36] I. Toda-Caraballo, J.S. Wrobel, S.L. Dudarev, D. Nguyen-Manh, P.E.J. Rivera-Daz-del Castillo, Interatomic spacing distribution in multicomponent alloys,
Acta Mater. 97 (2015) 156169.
[37] O.N. Senkov, J.M. Scott, S.V. Senkova, F. Meisenkothen, D.B. Miracle, C.F. Woodward, Microstructure and elevated temperature properties of a refractory
TaNbHfZrTi alloy, J. Mater. Sci. 47 (9) (2012) 40624074.
[38] G. Dirras, L. Lilensten, P. Djemia, M. Laurent-Brocq, D. Tingaud, J.P. Couzini,
L. Perrire, T. Chauveau, I. Guillot, Elastic and plastic properties of as-cast
equimolar TiHfZrTaNb high-entropy alloy, Mater. Sci. Eng. A 654 (2016) 3038.
[39] W.K.P. Hohenberg, Inhomogeneous electron gas, Phys. Rev. B 136 (3B) (1964)
864.
[40] W. Kohn, L.J. Sham, Self-consistent equations including exchange and correlation effects, Phys. Rev. 140A (1951) (1965) 1133.
[41] R.M. Dreizler, E.K.U. Gross, Density Functional Theory 1990.
[42] A. Zunger, S.H. Wei, L.G. Ferreira, J.E. Bernard, Special quasirandom structures,
Phys. Rev. Lett. 65 (3) (1990) 353356.
[43] S. Wei, L.G. Ferreira, J.E. Bernard, A. Zunger, Electronic properties of random
alloys: special quasirandom structures, Phys. Rev. B 42 (15) (1990) 9622.
[44] P. Soven, Coherent-potential model of substitutional disordered alloys, Phys.
Rev. 156 (3) (1967) 809813.
[45] B.L. Gyorffy, Coherent-potential approximation for a nonoverlapping-muntin-potential model of random substitutional alloys, Phys. Rev. B 5 (6) (1972)
23822384.
[46] D.W. Taylor, Vibrational properties of imperfect crystals, Phys. Rev. 156 (3)
(1967) 10171029.
[47] E. Pickering, N.G. Jones, High-entropy alloys: a critical assessment of their
founding principles and future prospects, Int. Mater. Rev. (2016) 120.

[48] O.N. Senkov, S.V. Senkova, C. Woodward, D.B. Miracle, Low-density, refractory
multi-principal element alloys of the CrNbTiVZr system: microstructure and
phase analysis, Acta Mater. 61 (5) (2013) 15451557.
[49] Y.D. Wu, Y.H. Cai, X.H. Chen, T. Wang, J.J. Si, L. Wang, Y.D. Wang, X.D. Hui, Phase
composition and solid solution strengthening effect in TiZrNbMoV high-entropy alloys, Mater. Des. 83 (2015) 651660.
[50] O.N. Senkov, S.V. Senkova, D.B. Miracle, C. Woodward, Mechanical properties of low-density, refractory multi-principal element alloys of the CrNbTiVZr
system, Mater. Sci. Eng. A 565 (2013) 5162.
[51] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made
simple, Phys. Rev. Lett. 77 (18) (1996) 3865.
[52] G. Kresse, J. Hafner, Ab initio molecular dynamics for liquid metals, Phys. Rev.
B 47 (1) (1993) 558561.
[53] G. Kresse, Ecient iterative schemes for ab initio total-energy calculations
using a plane-wave basis set, Phys. Rev. B 54 (16) (1996) 1116911186.
[54] G. Kresse, J. Furthmller, Eciency of ab-initio total energy calculations for
metals and semiconductors using a plane-wave basis set, Comput. Mater. Sci. 6
(1) (1996) 1550.
[55] L. Vitos, H.L. Skriver, B. Johansson, J. Kollr, Application of the exact mun-tin
orbitals theory: the spherical cell approximation, Comput. Mater. Sci. 18 (1)
(2000) 20.
[56] L. Vitos, I.A. Abrikosov, B. Johansson, Anisotropic lattice distortions in random
alloys from rst-principles theory, Phys. Rev. Lett. 87 (15) (2001) 156401.
[57] L. Vitos, Total-energy method based on the exact mun-tin orbitals theory,
Phys. Rev. B 64 (1) (2001) 14107.
[58] L. Vitos, Computational Quantum Mechanics for Materials Engineers: The
EMTO Method and Applications, London: Springer-Verlag. 2007.
[59] J. Kollr, L. Vitos, H. Skriver, From ASA towards the full potential, Lect. Notes
Phys. - Electron. Struct. Phys. Prop. Solids 535 (2000) 85114.
[60] L. Vitos, J. Kollr, H.L. Skriver, Full charge-density scheme with a kinetic-energy
correction: application to ground-state properties of the 4d metals, Phys. Rev.
B 55 (20) (1997) 1352113527.
[61] A. Van de Walle, M. Asta, G. Ceder, The alloy theoretic automated toolkit: a user
guide, Calphad 26 (4) (2002) 539553.
[62] A. van de Walle, Multicomponent multisublattice alloys, noncongurational
entropy and other additions to the alloy theoretic automated toolkit, Calphad
Comput. Coupling Phase Diagrams Thermochem. 33 (2) (2009) 266278.
[63] G. Kresse, From ultrasoft pseudopotentials to the projector augmented-wave
method, Phys. Rev. B 59 (3) (1999) 17581775.
[64] V.L. Moruzzi, J.F. Janak, K. Schwarz, Calculated thermal properties of metals,
Phys. Rev. B 37 (2) (1988) 790799.
[65] J. von Pezold, A. Dick, M. Frik, J. Neugebauer, Generation and performance of
special quasirandom structures for studying the elastic properties of random
alloys: application to Al-Ti, Phys. Rev. B 81 (9) (2010) 094203.
[66] F. Tasndi, M. Odn, I.A. Abrikosov, Ab initio elastic tensor of cubic Ti0.5 Al0.5 N
alloys: dependence of elastic constants on size and shape of the supercell
model and their convergence, Phys. Rev. B - Condens. Matter Mater. Phys. 85
(14) (2012) 19.
[67] L.-Y. Tian, Q.-M. Hu, R. Yang, J. Zhao, B. Johansson, L. Vitos, Elastic constants
of random solid solutions by SQS and CPA approaches: the case of fcc Ti-Al, J.
Phys: Condens. Matter 27 (31) (2015) 315702.
[68] A. Korzhavyi, Madelung energy for random metallic alloy in CPA, Phys. Rev. B
51 (9) (1995) 57735780.
[69] M.C. Gao, C. Niu, C. Jiang, D.L. Irving, Applications of Special Quasi-random Structures to High-entropy Alloys, High-Entropy Alloys, Springer. 2016,
333368.
[70] H. Levmki, M.P.J. Punkkinen, K. Kokko, L. Vitos, Flexibility of the quasi-non-uniform exchange-correlation approximation, Phys. Rev. B 89 (11) (2014)
115107.
[71] P. Sderlind, O. Eriksson, J.M. Wills, A.M. Boring, Theory of elastic constants of
cubic transition metals and alloys, Phys. Rev. B 48 (9) (1993) 58445851.
[72] I. Kunce, M. Polanski, J. Bystrzycki, Microstructure and hydrogen storage properties of a TiZrNbMoV high entropy alloy synthesized using Laser Engineered
Net Shaping (LENS), Int. J. Hydrogen Energy 39 (18) (2014) 99049910.
[73] F. Tian, L.K. Varga, N. Chen, J. Shen, L. Vitos, Empirical design of single phase
high-entropy alloys with high hardness, Intermetallics 58 (2015) 16.
[74] X. Yang, Y. Zhang, Prediction of high-entropy stabilized solid-solution in multi-component alloys, Mater. Chem. Phys. 132 (2-3) (2012) 233238.
[75] R. Ahuja, J.M. Wills, B. Johansson, O. Eriksson, Crystal structures of Ti, Zr, and Hf
under compression: theory, Phys. Rev. B 48 (22) (1993) 1626916279.
[76] M. Laurent-Brocq, L. Perrire, R. Pirs, Y. Champion, From high entropy alloys to
diluted multi-component alloys: range of existence of a solid-solution, Mater.
Des. 103 (2016) 8489.
[77] A.J. Zaddach, C. Niu, C.C. Koch, D.L. Irving, Mechanical properties and stacking fault energies of NiFeCrCoMn high-entropy alloy, JOM 65 (12) (2013)
17801789.
[78] J.W. Yeh, S.J. Lin, T.S. Chin, J.Y. Gan, S.K. Chen, T.T. Shun, C.H. Tsau, S.Y. Chou,
Formation of simple crystal structures in CuCoNiCrAlFeTiV alloys with multiprincipal metallic elements, Metall. Mater. Trans. A 35 (8) (2004) 25332536.

Please cite this article as: L-Y. Tian et al., Alloying effect on the elastic properties of refractory high-entropy alloys, Materials and Design
(2016), http://dx.doi.org/10.1016/j.matdes.2016.11.079

You might also like