Vol. 133 (2018) ACTA PHYSICA POLONICA A No.

Pressure Effects on Structural, Electronic, Elastic, and

Optical Properties of Cubic and Tetragonal Phases of BaZrO3
Xiong Yang, Ying Wang∗ , Qinggong Song, Yifei Chen and Yan Hong Xue
Institute of Low Dimensional Materials and Technology, College of Science, Civil Aviation University of China,
Tianjin 300300, China
(Received February 14, 2017; in final form May 24, 2017)
A recent experimental study found the phase transition from cubic to tetragonal phase at 17.2 GPa in BaZrO3 .
In this paper, we investigate the structural, electronic, elastic, and optical properties of BaZrO3 in the cubic
and tetragonal structures under pressure by using first-principle density-functional theory. The main results are
summarized as follows: (1) An indirect band structure is displayed for both cubic and tetragonal phases. The band
gap increases by ≈ 1 eV as a result of the phase transition. (2) The elastic constants Cij ’s indicate that the cubic
and tetragonal phases are mechanically stable. The easiest deformation mechanism in both phases is the shear
along the (100) plane. (3) The real and imaginary parts of dielectric function for the cubic and tetragonal phases
shift towards higher energies with pressure. Our results are in agreement with the available experimental and prior
theoretical data.
DOI: 10.12693/APhysPolA.133.1138
PACS/topics: pressure effect, electronic structure, elastic property, optical property

1. Introduction cubic and tetragonal phases under pressure. In this pa-

per, we explore how high pressure affects the structural,
Barium zirconate (BaZrO3 ) with a cubic perovskite electronic, elastic, and optical properties of BaZrO3 in
structure, a classic ABO3 compound, has been a popular the cubic and tetragonal phases by performing extensive
target because it displays a large number of desired phys- first-principle calculations. Our work provides a com-
ical properties, such as a high-temperature proton con- prehensive perspective into the pressure effects on the
ductor [1], a useful dielectric material for wireless commu- fundamental properties of BaZrO3 .
nications [2], a good substrate in thin-film deposition [3]
and solid oxide fuel cells [4]. The ABO3 crystals of cubic 2. Computational details
perovskite structure exhibit rich structural phase transi-
tions due to the changes of external conditions, such as The first-principle calculation has been successful in
pressure or temperature. For instance, Köhler et al. [5] accurate understanding of the structural stability, the
predicted theoretically a phase transition from cubic to physical properties of various compounds including per-
tetragonal structure in EuTiO3 , and subsequently veri- ovskite oxide, yttrium orthoaluminate, rhodium silicides,
fied in experiment that it occurs at T = 282 K. Recently, and carbon dioxide, under external conditions (pressure
high pressure experimental discoveries have shown that or impurity) [14–17]. Our first-principles calculations are
SrTiO3 and BaZrO3 undergo a structural phase transi- performed using density functional theory (DFT) within
tion from cubic to tetragonal phase, respectively [6, 7]. the plane-wave basis available in CASTEP code [18]. We
The phase transformation is generally considered to have used the generalized gradient approximation (GGA) in
the same driving force from symmetry breaking, i.e., the the scheme of Perdew–Wang 91 (PW91) [19]. The va-
phonon mode instability manifested clearly by imaginary lence electron configurations for Ba, Zr, O atoms are
frequency optical modes at the R-point in the phonon dealt with 2s2 2p4 , 5s2 5p6 6s2 and 4s2 4p6 4d2 5s2 , respec-
spectrum [8–13]. tively. The tolerances of stress, force and self-consistent
It is well known that the electronic, elastic, and optical field are 0.02 GPa, 0.01 eV/Å and 5.0 × 10−7 eV/atom,
properties of ABO3 compounds under external pressure respectively. Ultrasoft pseudopotentials within a 775 eV
will be altered with the phase transition, because of the cut-off energy are chosen to optimize structure. The k
change of the relative positions of atoms and interatomic mesh is sampled according to a Monkhorst–Pack scheme
distances. Such effects of pressure on the properties of with a spacing of 0.04 Å. Such parameters have been
compounds are not only attractive in physics, but also tested to be sufficient for convergence (< 1 meV).
intriguing in technology. However, to the best of our The phase transition from cubic to tetragonal phase
knowledge, there is no systematic research on the elec- was observed at 17.2 GPa in BaZrO3 [7]. We use the ex-
tronic, elastic, and optical properties of BaZrO3 in the perimental lattice constants a = 4.194 Å at 0 GPa in the
cubic phase, and a = 5.5074 Å, c = 7.9273 Å at 46.4 GPa
in the tetragonal phase [7]. As summarized in Tables I
and II, our optimized lattice parameters as well as in-
∗ corresponding author; e-mail: x-yang@[Link] ternal atomic positions agree well with the experimental
and other theoretical results.

(1138)
 Pressure Effects on Structural, Electronic. . . 1139

TABLE I
The relaxed structure lattice constants [Å], elastic
constants, bulk modulus B, shear modulus G and
the Young modulus E [GPa] of BaZrO3 in cubic
phase at 0 GPa compared with the experimental
data and prior theoretical results.

Source a C11 C12 C44 B G E

ours 4.223 295 82 85 153 93 260
expt. 4.192 - – – 189±26a 103B 243B
GGAc 4.207 322.5 79 70 157 91 229
LDAd 4.154 336 95 89.5 175.3 102 256
a Ref. [7], B Ref. [20], c Ref. [21], d Ref. [22]. Fig. 1. Pressure dependence of the unit cell volume in
the cubic phase, c/a and octahedral rotation degree θ in
TABLE II the tetragonal phase of BaZrO3 . The acceptable differ-
ence between the theoretical and experimental volumes
Lattice constants [Å] and atom positions in tetragonal in the cubic phase is due to the slightly exaggerated
phase of BaZrO3 at 46.4 GPa, computational and exper- lattice constant as a result of GGA method.
imental (Ref. [7]) data.
Source a c Ba Zr O) O(2) distinguish from those in Fig. 3. For both the phases,
ours 5.521 8.027 (0,0.5,0.25) (0,0,0) (0,0,0.25) (0.2988,0.7988,0) the top of the valence band mostly consists of O 2p and
expt. 5.507 7.927 (0,0.5,0.25) (0,0,0) (0,0,0.25) (0.2181,0.7131,0) Zr 4d electronic states, while Zr 4d states mainly domi-
nate the bottom of the conduction band. The energy gap
3. Results and discussion for the cubic phase is indirect along the R → Γ direction.
For the tetragonal phase, the minimum of the conduction
3.1. Structural properties band is at Γ , and the maximum of the valence band is at
The experimental studies confirmed that a phase tran- the high symmetry point (0,0.20,0) along the X (0,0.5,0)
sition from cubic (P m3̄m) to tetragonal (I4/mcm) phase to Γ direction. The energy at the point (0,0.20,0) is only
takes place in BaZrO3 under compression. The similar 0.045 eV higher than that at Γ , indicated by the arrow
transition behavior was observed in the other cubic per- in Fig. 3.
ovskites, such as SrTiO3 and EuTiO3 [5, 6]. It is sug-
gested that the cubic-tetragonal phase transition resulted
from the phonon mode instability with the rotation of
oxygen octahedra around [001] under high pressure [8–
13]. As shown in Fig. 1, our calculation presents the
pressure dependence of ZrO6 octahedra rotation angle in
BaZrO3 . The oxygen octahedra rotates by 6.396◦ (corre-
sponding to a Zr–O–Zr angle of 167.2◦ ) as a result of the
cubic-tetragonal phase transition at 17.2 GPa. The octa-
hedral rotation angle θ increases with increasing pressure,
and the extremely similar dependence of c/a on pressure
can be found. The change rate of rotation angle and Fig. 2. Electronic band structure (left part) and pro-
c/a decreases abruptly at 28 GPa, which implies that a jected density of states (right part) for the cubic phase
of BaZrO3 under 6 GPa.
displacive phase transition comes up at this pressure ac-
cording to the elastic properties shown in Fig. 5. Our
optimized tetragonal structure with ZrO6 octahedra tilt-
ing is only due to the rotations of ZrO6 octahedra along
the c direction, without any tilting along the a and b
directions, which is in agreement with the experimental
observation [7].
3.2. Electronic properties
The energy band structure and projected density of
states near the Fermi energy for the cubic and tetragonal
phases of BaZrO3 at 6 GPa and 35 GPa are shown in
Figs. 2 and 3. The overall band profiles at other pres- Fig. 3. Electronic band structure (left part) and pro-
sure for the two phases are found to be similar to Figs. 2 jected density of states (right part) for the tetragonal
phase of BaZrO3 under 35 GPa.
and 3. The Brillouin zone shape of the cubic phase dif-
fers from that of the tetragonal phase, hence the selected Figure 4 indicates the pressure-induced band gap vari-
high symmetry points along the transverse axis in Fig. 2 ation for BaZrO3 . Our calculated band gap at equilib-
1140 Xiong Yang et al.

15
= 4 (S11 + S22 + S33 ) − 4 (S12 + S23 + S31 )
GR
+3 (S44 + S55 + S66 ) , (4)
where Cij and Sij describe the elastic stiffness and the
elastic compliance, respectively. In addition, the so-
called Hill theory [28] gives
B = (BV + BR ) /2, (5)
G = (GV + GR ) /2. (6)
the Young modulus (E) and the Poisson ratio (γ) are
Fig. 4. Band gap of the cubic and tetragonal phases of obtained by the relation
BaZrO3 as a function of pressure.
E = 9BG/(3B + G), (7)
rium is 3.164 eV, which can be compared to the the- γ = (3B − 2G) / [2(3B + G)] . (8)
oretical value of 3.23 eV [21]. The theoretical values The elasticity of a cubic crystal is characterized by three
using GGA method are underestimated with respect to independent modulus parameters: C11 , C12 and C44 .
the experimental value of 5.33 eV [23]. Moreover, GGA The symmetry of the tetragonal phase is below that of
usually agrees with the more time-consuming Heyd– the cubic phase, hence the independent elastic constants
Scuseria–Ernzerhof (HSE06) on pressure coefficients of increase to C11 , C12 , C13 , C33 , C44 and C66 . The cal-
band gap [24–26]. Also, the different exchange correla- culated elastic constants of BaZrO3 at equilibrium are
tion functionals appear not to change the trends in the represented in Table I. Our results at zero pressure are
dependence of the band gap on pressure. Thus, pressure in good agreement with previous calculations and exper-
coefficients of the band gap computed using DFT could imental data. Figure 5 shows the pressure effects on the
be observed in experiment. As shown in Fig. 4, the band elastic constants in the cubic and tetragonal phases. The
gap energy increases by ≈ 1 eV while phase transition cubic phase within 0–17.2 GPa and the tetragonal phase
from cubic to tetragonal occurring at 17.2 GPa, because within 17.2–46.4 GPa are mechanically stable, because
of the decrease of conduction band width. In both the the Cij matrix satisfies the well-known Born stability cri-
phases, the pressure results in the linear increase of the teria [29]. For cubic phase, they are
band gap. This pressure effect is very similar to that of C11 > 0, C11 > C12 , C44 > 0. (9)
SrTiO3 in Ref. [27]. The change rate of the band gap For tetragonal phase, they are
for the tetragonal phase drops gradually with increase
C11 > 0, C33 > 0, C44 > 0, C66 > 0, C11 > C12 ,
of pressure. We note that while compression increases
from 17.2 GPa (0 GPa) to 46.4 GPa (17.2 GPa) in the C11 + C33 − 2C13 > 0,
tetragonal (cubic) phase, the band gap increases only by 2(C11 + C12 ) + C33 + 4C 13 > 0. (10)
4.8% (6.3%). The cubic elastic coefficients C11 , C12 and the tetrago-
nal coefficients C11 , C12 , C13 , C33 increase with pressure,
3.3. Elastic properties
while the cubic C44 and the tetragonal C44 , C66 do not
The elastic constants describe the nature character of vary much with pressure. The curves of cubic C11 , C12
the forces operating in solids and provide information and tetragonal C11 , C12 , C13 , C33 show a leap at about
on the stiffness and stability of materials. Based on the 12 GPa and 28 GPa respectively, which indicates that a
fully structure relaxation under pressure in the cubic and displacive phase transition occurs at these pressures. The
tetragonal phases, the elastic tensor Cij is calculated similar feature has been observed in the pressure effects
by using the “volume-conserving technique”. We define of the elastic constants of SrTiO3 in Ref. [30]. The C44
the bulk modulus and shear modulus by the Reuss the- constant is almost unchanged in both cubic and tetrago-
ory (BR , GR ) and the Voigt theory (BV , GV ) respec- nal phases, implying that the shear along the (100) plane
tively [28]. In the commonly used Voigt approximation, is the easiest and does not depend on the phase. The
1 tetragonal C11 is much smaller than C33 , which means
BV = [(C11 + C22 + C33 ) + 2(C12 + C23 + C31 )] , that a and b axes in this phase are easier to be com-
9
(1) pressed than in the cubic one.
1 Based on the obtained elastic coefficients of BaZrO3 ,
GV = [(C11 + C22 + C33 ) − (C12 + C23 + C31 )
15 we further predict its noteworthy elastic properties in the
+3 (C44 + C55 + C66 )] . (2) cubic and tetragonal phases, such as bulk modulus, shear
modulus, the Young modulus, the Poisson ratio, and
On the other hand, according to the Reuss approxima-
anisotropy factor. The calculated results are summarized
tion,
in Tables III and IV. It can be seen that the three mod-
1
= (S11 + S22 + S33 ) + 2(S12 + S23 + S31 ), (3) uli increase as pressure increases, and the Young modulus
BR E is always largest, shear modulus G smallest. It is well
 Pressure Effects on Structural, Electronic. . . 1141

TABLE III index AU to estimate the anisotropic characteristic,

Bulk modulus B, shear molulus G, the Young modulus which is as follows:
E, ([GPa]), the Poisson ratio γ, and anisotropy factor 5GV BV
AU = + − 6. (11)
AU of BaZrO3 in cubic phase at several pressures. GR BR
For an isotropic crystal, AU = 0. The departure of
P B G E B/G γ AU
AU from zero defines the degree of elastic anisotropy.
0 153 93 260 1.645 0.217 0.063
The calculated results of AU in Tables III and IV devi-
3 165 98 282 1.684 0.214 0.102
ate slightly from 0, indicating the relatively small elas-
6 187 102 308 1.833 0.225 0.145 tic anisotropy characteristic of BaZrO3 in the cubic and
9 203 107 331 1.897 0.228 0.187 tetragonal phases. It is notable that the increasing pres-
12 197 111 346 1.775 0.207 0.231 sure leads to more and more heavy extent of anisotropy
15 208 115 368 1.809 0.206 0.278 for both cubic and tetragonal phases, except for the leap
17.2 227 118 386 1.924 0.216 0.308 at about 25 GPa in the tetragonal phase.

TABLE IV
Bulk modulus B, shear molulus G, the Young modulus
E, ([GPa]), the Poisson ratio γ, and anisotropy factor
AU of BaZrO3 in tetragonal phase at several pressures.

P B G E B/G γ AU
17.2 223 115 294 1.939 0.280 0.212
20 234 118 303 1.983 0.284 0.234
22.5 243 120 309 2.025 0.288 0.263
25 251 120 310 2.092 0.294 0.2344
30 265 118 308 2.246 0.306 0.2556
Fig. 5. Elastic constants of the cubic and tetragonal
35 287 124 325 2.315 0.311 0.297 phases of BaZrO3 as a function of pressure.
40 305 126 332 2.420 0.318 0.321
46.4 326 128 340 2.547 0.326 0.357 3.4. Optical properties
The dielectric function ε(ω) can be used to describe
the linear response of the system to electromagnetic ra-
known that the Young modulus is a crucial parameter for diation, which is connected with the interaction between
selecting materials in engineering design. The material photons and electrons [34]. The imaginary part ε2 (ω)
is stiffer for the larger the Young modulus. Our calcu- can be calculated from the momentum matrix elements
lations indicate that stiffness of BaZrO3 becomes higher between the occupied and unoccupied electronic states
and higher with increase of pressure for both the cubic within selection rules, and the real part ε1 (ω) is linked
and tetragonal phases. the Poisson ratio γ can describe with ε2 (ω) by the Kramer–Krönig relations. The other
the characteristic of the bonding forces to some extent. linear optical parameters, such as refractive index and
γ = 0.5 is the lower limit for central-force solids and 0.5 electron energy-loss coefficient, can be evaluated on the
is the upper limit, which corresponds to infinite elastic basis of ε1 (ω) and ε2 (ω). To check the calculation relia-
anisotropy [31]. The obtained values of the cubic phase bility with using GGA-PW91 and ultrasoft pseudopoten-
within the whole range of 0–17.2 GPa are below 0.25, tial scheme, we also calculated ε1 (ω) and ε2 (ω) of the cu-
however, the values of the tetragonal phase are above bic SrTiO3 with perovskite structure similar to BaZrO3 ,
0.25 up to 46.4 GPa. This indicates that the interatomic which agree well with the experimental data, as shown in
forces of BaZrO3 are non-central for the cubic phase, but Fig. 6. In our calculations of optical spectra, a sufficiently
central for the tetragonal phase. In order to evaluate the dense k-mesh is applied by the separation of optical k-
ductile/brittle nature of materials, Pugh [32] proposed a points 0.02 1/Å.
criterion. When the ratio of bulk to shear modulus B/G The real and imaginary parts of dielectric function of
is bigger (less) than 1.75, the material is considered duc- BaZrO3 in the cubic and tetragonal phases under pres-
tile (brittle). From the predicted values of B/G, it can sure are displayed in Figs. 7 and 8. Since the tetragonal
be judged that BaZrO3 in the cubic phase is brittle below phase of BaZrO3 is optically anisotropic, components of
6 GPa, ductile above 6 GPa. The ductility of BaZrO3 in ε1 (ω) and ε2 (ω) corresponding to the polarization vec-
the tetragonal phase is more and more significant with tors perpendicular (E⊥c) and parallel (E k c) to the
increase of pressure. crystallographic c axis have been considered. Figures 7
The evaluation of elastic anisotropy in material is of and 8 only shows the curves with E⊥c, because the curve
great significance to understand the mechanical prop- structures of E⊥c and E k c are similar to each other.
erties of crystal. An appropriate measure to quantify The whole ε1 (ω) and ε2 (ω) curves for two phases shift
the degree of anisotropy is necessary. Ranganathan and towards higher energies with increase of pressure, show-
Ostoja-Starzewski [33] developed a universal anisotropy ing blueshift compared to the equilibrium state. The first
1142 Xiong Yang et al.

Fig. 6. Real and imaginary parts of dielectric func- Fig. 9. Static dielectric constant of the cubic and
tion ε(ω) of SrTiO3 . The experimental data are from tetragonal phases of BaZrO3 as a function of pressure.
Ref. [35].

Fig. 10. Refractive index of the cubic and tetragonal

phases of BaZrO3 as a function of pressure.
Fig. 7. Real and imaginary parts of dielectric function
ε(ω) of the cubic phase of BaZrO3 as a function of pres-
sure. larization is smaller than that of the extraordinary po-
larization for the tetragonal phase (right part of Fig. 10),
peak of ε2 (ω) originates from the transfer of the highest and BaZrO3 compound is positive uniaxial crystal.
valence electrons (O p states) to the lowest conduction
bands (Zr d states). The shift of the first peak implies the 4. Conclusions
enlargement of the band gap resulted by pressure, which
is consistent with Fig. 4. In terms of the low-frequency In this paper, we present comprehensive first-principle
behavior of the dielectric function, the static dielectric calculation results about the pressure effects on the fun-
constant ε1 (0) and zero frequency refractive index n(0) damental properties of BaZrO3 in the cubic and tetrag-
are of interest in experiment. onal structures for the first time. The band gap for the
As displayed in Figs. 9 and 10, the transition from cubic and tetragonal phases is indirect, and it increases
cubic to tetragonal phase reduces the values of static di- by ≈ 1 eV while the phase transition occurs. The elas-
electric constant ε1 (0) and refractive index n(0). The tic constants Cij ’s indicate that the cubic and tetragonal
pressure effects lead to the decrease (increase) of ε1 (0) phases are stable. The easiest deformation mechanism in
and n(0) for the cubic (tetragonal) phase, however, the both phases is the shear along the (100) plane. The real
drop (raise) is rather small compared to the initial value. and imaginary parts of dielectric function for the cubic
It is worthwhile to note that the n(0) of the ordinary po- and tetragonal phases shift towards higher energies with
pressure. The pressure effects on the other elastic and
optical parameters, including bulk modulus, shear mod-
ulus, the Young modulus and refractive index are also
presented.

Acknowledgments

This work was supported by the Fundamental Re-

search Funds for the Central Universities (3122014K003).

References

[1] K.D. Kreuer, Annu. Rev. Mater. Res. 33, 333 (2003).
Fig. 8. Real and imaginary parts of dielectric function
ε(ω) of the tetragonal phase of BaZrO3 as a function of [2] M.T. Sebastian, Dielectric Materials for Wireless
pressure. Communication, Elsevier, Amsterdam 2008.
 Pressure Effects on Structural, Electronic. . . 1143

[3] P.S. Dobal, A. Dixit, R.S. Katiyar, Z. Yu, R. Guo, [18] V. Milman, B. Winkler, J.A. White, C.J. Packard,
A.S. Bhalla, J. Appl. Phys. 89, 8085 (2001). M.C. Payne, E.V. Akhmatskaya, R.H. Nobes,
[4] M. Miura, B. Maiorov, T. Kato, T. Shimode, Int. J. Quantum Chem. 77, 895 (2000).
K. Wada, S. Adachi, K. Tanabe, Nat. Commun. 4, [19] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett.
2499 (2013). 77, 3865 (1996).
[5] J. Köhler, R. Dinnebier, A. Bussmann-Holder, Phase [20] S. Yamanaka, H. Fujikane, T. Hamaguchi, H. Muta,
Transit. 85, 949 (2012). T. Oyama, T. Matsuda, S. Kobayashi, K. Kurosaki,
[6] M. Guennou, P. Bouvier, B. Krikler, J. Kreisel, J. Alloys Comp. 359, 109 (2003).
D. Machon, Phys. Rev. B 81, 054115 (2010). [21] R. Terki, H. Feraoun, G. Bertrand, H. Aourag,
[7] X. Yang, Q.J. Li, R. Liu, B. Liu, H.F. Zhang, Phys. Status Solidi B 242, 1054 (2005).
S.Q. Jiang, J. Liu, B. Zou, T. Cui, B.B. Liu, [22] R.D. King-Smith, D. Vanderbilt, Phys. Rev. B 49,
J. Appl. Phys. 115, 124907 (2014). 5828 (1994).
[8] K.Z. Rushchanskii, N.A. Spaldin, M. Lezaic, [23] J. Robertson, J. Vac. Sci. Technol. B 18, 1785 (2000).
Phys. Rev. B 85, 104109 (2012). [24] L. Qin, Y. Duan, H. Shi, L. Shi, G. Tang,
[9] J.W. Bennett, I. Grinberg, A.M. Rappe, Phys. Rev. B J. Phys. Condens. Matter 25, 045801 (2013).
73, 180102 (2006). [25] Yifeng Duan, Lixia Qin, Liwei Shi, Gang Tang,
[10] A. Bilic, J.D. Gale, Phys. Rev. B 79, 174107 (2009). Hongliang Shi, Appl. Phys. Lett. 100, 022104 (2012).
[11] A.I. Lebedev, I.A. Sluchinskaya, Phys. Solid State 55, [26] Dong Lv, Yifeng Duan, Botao Zhao, Lixia Qin, Liwei
1941 (2013). Shi, Gang Tang, Hongliang Shi, J. Appl. Phys. 114,
[12] A.R. Akbarzadeh, I. Kornev, C. Malibert, L. Bel- 023514 (2013).
laiche, J.M. Kiat, Phys. Rev. B 72, 205104 (2005). [27] L. Khaber, A. Beniaiche, A. Hachemi, Solid State
[13] W. Zhong, D. Vanderbilt, Phys. Rev. Lett. 74, 2587 Commun. 189, 32 (2014).
(1995). [28] R. Hill, Proc. Phys. Soc. Lond. A 65, 349 (1952).
[14] Cheng Lu, Jingjing Wang, Ping Wang, Xinxin [29] M. Born, K. Huang, Dynamical Theory of Crystal Lat-
Xia, Yuanyuan Jin, Peifang Lie, Gang Bao, tices, Oxford University Press, New York 1988.
Phys. Chem. Chem. Phys. 19, 1420 (2017). [30] A. Hachemi, H. Hachemi, A. Ferhat-Hamida,
[15] Meng Ju, Cheng Lu, Yauyuen Yeung, Xiaoyu L. Louail, Phys. Scr. 82, 025602 (2010).
Kuang, Jingjing Wang, Yongsheng Zhu, ACS [31] P. Ravindran, L. Fast, P.A. Korzhavyi, B. Johansson,
Appl. Mater. Inter. 8, 30422 (2016). J. Wills, O. Eriksson, J. Appl. Phys. 84, 4891 (1998).
[16] Jing-Jing Wang, A. Hermann, Xiao-Yu Kuang, Yuan- [32] S.F. Pugh, Philos. Mag. 45, 823 (1954).
Yuan Jin, Cheng Lu, Chuan-Zhao Zhang, Meng Ju,
Meng-Ting Sic, Toshiaki Iitaka, RSC Adv. 5, 53497 [33] S.I. Ranganathan, M. Ostoja-Starzewski,
(2015). Phys. Rev. Lett. 101, 055504 (2008).
[17] Cheng Lu, Maosheng Miao, Yanming Ma, [34] Jian Sun, Hui-Tian Wang, Nai-Ben Ming, Julong He,
J. Am. Chem. Soc. 135, 14167 (2013). Yongjun Tian, Appl. Phys. Lett. 84, 4544 (2004).
[35] M. Cardona, Phys. Rev. 140A, 651 (1965).