Hindawi

Advances in Materials Science and Engineering

Volume 2017, Article ID 3981317, 13 pages
https://doi.org/10.1155/2017/3981317

Research Article
Consistent Atomic Geometries and Electronic Structure of Five
Phases of Potassium Niobate from Density-Functional Theory

Falko Schmidt,1 Marc Landmann,1 Eva Rauls,1 Nicola Argiolas,2 Simone Sanna,1
W. G. Schmidt,1 and Arno Schindlmayr1
1
Department Physik, Universität Paderborn, 33095 Paderborn, Germany
2
Dipartimento di Fisica e Astronomia, Universitá di Padova, 35131 Padova, Italy

Correspondence should be addressed to Arno Schindlmayr; [email protected]

Received 8 September 2016; Revised 11 November 2016; Accepted 24 November 2016; Published 30 January 2017

Academic Editor: Pascal Roussel

Copyright © 2017 Falko Schmidt et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

We perform a comprehensive theoretical study of the structural and electronic properties of potassium niobate (KNbO3 ) in
the cubic, tetragonal, orthorhombic, monoclinic, and rhombohedral phase, based on density-functional theory. The influence
of different parametrizations of the exchange-correlation functional on the investigated properties is analyzed in detail, and the
results are compared to available experimental data. We argue that the PBEsol and AM05 generalized gradient approximations as
well as the RTPSS meta-generalized gradient approximation yield consistently accurate structural data for both the external and
internal degrees of freedom and are overall superior to the local-density approximation or other conventional generalized gradient
approximations for the structural characterization of KNbO3 . Band-structure calculations using a HSE-type hybrid functional
further indicate significant near degeneracies of band-edge states in all phases which are expected to be relevant for the optical
response of the material.

1. Introduction recently discovered in KNbO3 nanomaterials synthesized via

a hydrothermal method [7]. It has long been known that
The study of perovskites, oxidic compounds with the chem- the room-temperature orthorhombic phase, in particular, is
ical formula ABO3 and an ideal or distorted cubic crystal characterized by unusual dielectric properties and an excep-
structure, is of both academic and technological interest. tionally large nonlinear optical coefficient. For this reason,
On one hand, perovskite minerals are the most abundant KNbO3 is widely used in nonlinear optics and holographic
components in the Earth’s mantle [1]. On the other hand, they applications, where it represents a viable alternative to the
exhibit a large variety of interesting physical properties [2, 3], more diffused LiNbO3 .
which can be modified by composition-induced structural Due to its technological importance, pure and defec-
modifications. tive KNbO3 has been subject to many experimental [8–
Among the human-made perovskites, potassium niobate 12] and theoretical [13–19] studies. The previous theoretical
(KNbO3 ) belongs to the important class of perovskite- investigations were typically aimed at the description of the
structure ferroelectrics [4]. It crystallizes at about 1325 K in ground-state structure and the identification of defect centers
a paraelectric cubic phase (space group 𝑃𝑚3𝑚) and, similar relevant for the optical properties. For the most part, they
to BaTiO3 [5], undergoes three successive ferroelectric phase used density-functional theory [20, 21] with classical local or
transitions [6]: to a tetragonal phase (space group 𝑃4𝑚𝑚) semilocal exchange-correlation functionals or even simpler
at 691 K, from the tetragonal to an orthorhombic phase semiempirical models. While these approaches are appro-
(space group 𝐴𝑚𝑚2) at 498 K, and from the orthorhombic priate for atomic geometries and the mechanical properties
to a rhombohedral phase (space group 𝑅3𝑚) at 263 K. In of a material if not too high accuracy is required, they
addition, a new monoclinic phase (space group 𝑃𝑚) was very are much less accurate with regard to the electronic and
2 Advances in Materials Science and Engineering

optical properties, because important factors like quasipar- test several different schemes in order to compare their per-
ticle renormalization of the electronic band gap, thermal formance, namely, the local-density approximation (LDA),
effects due to the electron-phonon coupling, or excitonic semilocal generalized gradient approximations (GGA), a
effects in the linear and nonlinear optical absorption are semilocal meta-GGA, and a nonlocal hybrid functional.
missing. Furthermore, most of the published work so far In detail, we use the LDA as parametrized by Perdew
concentrated on the structurally simple cubic and tetragonal and Zunger [31]. The conventional Perdew-Burke-Ernzerhof
phases, while less attention has been given to the remaining (PBE) parametrization [32] and the more recently introduced
polymorphs. As a consequence, a thorough quantitative ab PBEsol [33] and AM05 [34] functionals are considered on the
initio study of the electronic and optical properties of KNbO3 , GGA side; the latter two are chosen because they are known
especially of the more complex but technologically relevant to yield accurate lattice parameters comparable to hybrid-
room-temperature phase, with state-of-the-art methods that functional DFT calculations at a moderate computational
go beyond density-functional theory and take the above- cost [35]. Furthermore, the revised Tao-Perdew-Staroverov-
mentioned effects into account, as recently carried out for its Scuseria (RTPSS) meta-GGA [36] is employed. Band-
sister material LiNbO3 [22–25], is still missing. The accurate structure calculations are additionally performed for the
theoretical description of the atomic structure of the relevant relaxed PBEsol geometries with the Heyd-Scuseria-Ernzer-
KNbO3 phases remains a prerequisite for all such studies, hof (HSE) screened-Coulomb-potential hybrid functional
however. [37, 38]. The fraction of the exact exchange from Hartree-
In order to pave the way for future investigations along Fock theory, 25% in the conventional HSE06 functional, is
these lines, here we perform a comprehensive study of the increased to 30% to match the experimentally observed band
structural and electronic properties of all the five known gap of the cubic KNbO3 phase. Nonlocal HSE-type hybrid
phases of KNbO3 , including the newly discovered monoclinic functionals offer a reliable way to approximate more rigorous
phase, based on density-functional theory. The influence band-structure calculation schemes within the framework of
of different parametrizations of the exchange-correlation quasiparticle theory and thus effectively overcome the band
functional on the investigated properties is discussed in gap problem of conventional DFT calculations with local or
detail, and the results are compared to available experimental semilocal exchange-correlation functionals at a lower numer-
data. Such a systematic approach is especially important ical cost [39, 40].
for KNbO3 because the energetic separation and the struc- The electronic wave functions are expanded into plane
tural differences between the various phases are small and waves up to a kinetic energy of 600 eV for the lattice-para-
therefore potentially sensitive to details of the computational meter optimization. All structural degrees of freedom are
procedure. Indeed, our results show that several recently relaxed until the forces on each atom are below 0.001 meV/Å.
introduced exchange-correlation functionals give, overall, a The cutoff energy is reduced to 400 eV in the hybrid-func-
better description of the atomic geometries throughout the tional band-structure calculations. The Brillouin-zone inte-
various phases than common standard functionals as used in grations are performed using a shifted Monkhorst-Pack 6 × 6
earlier studies. Band-structure calculations are additionally × 6 k-point mesh for all KNbO3 phases. In calculations with
performed with a nonlocal hybrid functional to correct the the HSE hybrid functional, a regular Γ-centered 6 × 6 × 6 k-
well-known underestimation of the fundamental band gap point mesh is used instead.
incurred with conventional local or semilocal functionals,
which stems from the absence of a derivative discontinuity 3. Results and Discussion
with respect to the particle number as in the exact functional
[26, 27]. Figure 1 illustrates the five different phases of KNbO3 con-
This paper is organized as follows. In Section 2, we give sidered here with exaggerated deformations from the cubic
an overview of our computational method. In Section 3, configuration (Figure 1(a)). Starting from the cubic phase,
we present our results for the optimized atomic geometries the tetragonal phase is reached by elongation along 𝑎 axis
of the five considered phases of KNbO3 and compare the and compression along the other two axes (Figure 1(b)).
performances of different exchange-correlation functionals Further distortions of the lattice parameters transform the
in relation to experimental data. Furthermore, we discuss the tetragonal phase into an orthorhombic polymorph with
effects of the differently broken symmetries on the electronic distinct lattice parameters 𝑎󸀠 , 𝑏󸀠 , and 𝑐󸀠 (Figure 1(c), left). The
band structure through the successive phase transitions. We crystal structure of orthorhombic KNbO3 , whose rectangular
finally summarize our conclusions in Section 4. unit cell contains two formula units, that is, ten atoms, can
alternatively be characterized by a smaller rhombic unit cell
2. Computational Method with lattice parameters 𝑎 ≠ 𝑏 = 𝑐 and 𝛼 ≠ 𝛽 = 𝛾 = 90∘
which contains only one formula unit (Figure 1(c), right);
All calculations of this study are performed with the Vienna it is this nonrectangular primitive unit cell that we employ
Ab initio Simulation Package (VASP) [28, 29], a plane- in all structure optimizations and electronic-structure cal-
wave implementation of density-functional theory (DFT). culations. The relation between the orthorhombic and the
The electron-ion interaction is described by the projector- rhombic unit cells, including the ion displacements, is further
augmented-wave (PAW) scheme [29, 30], where the K 3𝑠 and illustrated in Figure 2. The orthorhombic phase transforms
the Nb 4𝑠 electrons are treated explicitly as valence states. into the monoclinic phase by choosing mutually distinct
For the exchange and correlation between the electrons, we lattice parameters 𝑎 ≠ 𝑏 ≠ 𝑐 (Figure 1(d)). Alternatively,
Advances in Materials Science and Engineering 3

b c b
a 
b b
a
a a a
a

K a=b=c K a ≠ b = c K a  = ̸ b  = ̸ c a ≠ b = c
∘
Nb  =  =  = 90 Nb  =  =  = 90∘ Nb  =  =  = 90  ≠  =  = 90∘
O Pm3m O P4mm O Amm 2
(a) (b) (c)

c a
b a

a a

K a ≠ b ≠ c K a=b=c
Nb  ≠  =  = 90∘ Nb  =  =  ≠ 90∘
O Pm O R3m
(d) (e)

Figure 1: Different phases of KNbO3 : (a) cubic, (b) tetragonal, (c) orthorhombic (left: conventional orthorhombic unit cell, right: primitive
rhombic unit cell), (d) monoclinic, and (e) rhombohedral with lattice parameters, angles, and space groups in Hermann-Mauguin notation.
The deformations are actually small and are exaggerated here for clarity.

the rhombohedral phase is obtained by choosing equal lattice x=0 x = a/2

parameters 𝑎 = 𝑏 = 𝑐 together with equal but acute angles
𝛼 = 𝛽 = 𝛾 (Figure 1(e)).
c
3.1. Structure Optimization. In order to study the influence c
of different approximations for exchange and correlation on
the DFT results for KNbO3 , we perform complete structure c b
c b
optimizations of the lattice parameters and the atomic dis-
placements for all five phases using the LDA, PBE, PBEsol, b
b
AM05, and RTPSS functionals. We start with the high- z
temperature cubic phase and then follow the order in which
the phases would occur for decreasing temperature. x y
Table 1 lists the calculated lattice parameter 𝑎 and the
K
optimized unit-cell volume 𝑉 for the cubic phase obtained
Nb
with the different exchange-correlation functionals as well O
as experimental data for comparison. Consistent with the
cubic symmetry, all lattice parameters have precisely the same Figure 2: Atomic displacements in the conventional orthorhombic
(solid lines) and primitive rhombic (dashed lines) unit cells of
value, and all lattice vectors are perpendicular to each other.
the orthorhombic KNbO3 polymorph. The unit cell is split into
There are no internal degrees of freedom and hence no atomic two layers of atoms perpendicular to 𝑎 axis. The deformations are
displacements in the cubic phase. At first glance, the PBE exaggerated for clarity.
results appear to agree best with the experimental lattice
constant, but it must be borne in mind that the measurement
was performed at 708 K, whereas the calculations correspond
to 0 K. The good quantitative agreement is in fact fortuitous at zero temperature. The LDA predicts the smallest value for
and due to an error cancellation between the neglect of 𝑎, while the results from other functionals are in between;
thermal expansion and the well-known tendency of PBE (and the same applies to the equilibrium unit-cell volume. In
other conventional GGAs) to overestimate lattice constants concordance with literature data on various other materials,
4 Advances in Materials Science and Engineering

Table 1: Lattice parameter 𝑎 and equilibrium unit-cell volume 𝑉 calculated with different exchange-correlation functionals for the cubic
phase of KNbO3 . Data in brackets represent relative deviations from the experimental values in percent.

𝑃𝑚3𝑚 LDA PBE PBEsol AM05 RTPSS Expt.a

𝑎 (Å) 3.954 (−1.69) 4.024 (+0.05) 3.985 (−0.92) 3.989 (−0.82) 4.008 (−0.35) 4.022
3
𝑉 (Å ) 61.72 (−5.13) 65.16 (+0.15) 63.28 (−2.74) 63.47 (−2.44) 64.38 (−1.05) 65.06
a
Ref. [41]; measured at 708 K.

Table 2: Lattice parameters 𝑎 and 𝑏, ratio 𝑎/𝑏, equilibrium unit-cell volume 𝑉, atomic displacements along 𝑎 axis in percent of 𝑎, and
total-energy difference Δ𝐸 per formula unit with respect to cubic KNbO3 calculated with different exchange-correlation functionals for the
tetragonal phase of KNbO3 . Data in brackets represent relative deviations from the experimental values in percent.

P4mm LDA PBE PBEsol AM05 RTPSS Expt.a

𝑎 (Å) 3.989 (−1.82) 4.193 (+3.20) 4.058 (−0.12) 4.098 (+0.86) 4.085 (+0.54) 4.063
𝑏 (Å) 3.945 (−1.30) 3.994 (−0.08) 3.969 (−0.70) 3.968 (−0.73) 3.992 (−0.13) 3.997
𝑎/𝑏 1.0112 (−0.52) 1.0498 (+3.28) 1.0224 (+0.58) 1.0328 (+1.60) 1.0233 (+0.67) 1.0165
3
𝑉 (Å ) 62.08 (−4.36) 66.87 (+3.02) 63.93 (−1.51) 64.52 (−0.60) 65.10 (+0.29) 64.91
ΔK 1.072 1.762 1.462 1.635 1.634 1.8
ΔNb 0 0 0 0 0 0
ΔO1 2.859 6.142 4.109 4.972 4.277 4.4
ΔO2 3.026 4.962 3.850 4.352 3.934 4.0
ΔO3 3.026 4.962 3.850 4.352 3.934 4.0
Δ𝐸 (eV) −0.0069 −0.0522 −0.0201 −0.0330 −0.0231
a
Ref. [41]; measured at 543 K.

we find that PBEsol, AM05, and RTPSS provide accurate and clearly shows the best performance in the description of these
quite similar lattice parameters [36, 42–46]. internal degrees of freedom, aside from the accidentally close
For the tetragonal phase, the data listed in Table 2 is PBE displacement of the K atom. Overall, we thus conclude
obtained with equal lattice parameters 𝑏 and 𝑐 and perpen- that PBEsol and RTPSS are well suited to describe the tetrag-
dicular lattice vectors. In contrast to the cubic phase, the onal phase of KNbO3 , although RTPSS provides the most
displacements do not vanish in this case. If we keep the Nb balanced description of all structural degrees of freedom. This
atom fixed at its ideal position in the center of the unit cell, highlights the importance of incorporating all external as well
we observe that all other atoms shift towards positive values as internal degrees of freedom in the evaluation of functional
along 𝑎 axis; displacements along the other axes are forbidden performance for structural properties.
due to symmetry constraints. Since the experimental lattice Table 3 contains our results for the orthorhombic phase.
parameters shown for comparison refer to 543 K and include We point out again that the primitive rhombic unit cell is not
effects of thermal expansion, the ratio 𝑎/𝑏 is also listed. The rectangular, as only the double unit cell exhibits the actual
lattice parameter 𝑎 is best matched by PBEsol, while the PBE orthorhombic structure. The arrows in Figure 2 indicate the
value for 𝑏 is closest to the experiment. The latter is fortuitous, direction of the displacements in the plane perpendicular to
however, as indicated by the large overestimation of 𝑎 which 𝑎 axis; there are no detectable displacements parallel to 𝑎.
exceeds 3% and constitutes the largest deviation among all As can be seen, the atoms shift mainly along 𝑐󸀠 direction
tested functionals. The deficiency of the PBE functional is of the double unit cell. In the rotated basis system of the
further reflected in the ratio 𝑎/𝑏 of 1.0498, which shows the primitive single unit cell, the displacements have roughly
strongest deviation from the experimental value of 1.0165, equal components along 𝑏 and 𝑐 directions. As before, the
while LDA (1.0112) slightly underestimates the experimental LDA lattice parameters and volume substantially underesti-
value, PBEsol (1.0224) and RTPSS (1.0233) slightly overes- mate the experimental data. In addition, LDA tends strongly
timate it by a similar amount, and AM05 (1.0328) yields a towards cubic symmetry, as indicated by the almost right
somewhat larger ratio. Despite the reasonably accurate LDA angle of 89.91∘ , 𝑎/𝑏 ratio of 0.9922 that is close to unity, and
lattice-parameter ratio, the associated equilibrium volume of the significantly underestimated atomic displacements. The
3
the unit cell of 62.08 Å shows the poorest agreement with PBE parametrization shows the smallest (for 𝑎) and largest
3 (for 𝑏) deviations from the experimental values among all
the experimental volume of 64.91 Å . The equilibrium volume considered functionals, indicating a similar imbalance in the
3
is excellently reproduced by the RTPSS value of 65.10 Å , description of lattice parameters as observed for the tetrag-
followed by the AM05 value of 64.52 Å. As for the relative onal phase. In contrast, the functionals PBEsol, AM05, and
atomic displacements, all functionals except the LDA, which RTPSS provide reasonable lattice parameters with compara-
predicts systematically too small shifts, are in reasonable bly small deviations from the experimental values for both 𝑎
agreement with the experimental data. However, RTPSS and 𝑏. This more balanced description of lattice parameters
Advances in Materials Science and Engineering 5

Table 3: Lattice parameters 𝑎 and 𝑏, angle 𝛼, ratio 𝑎/𝑏, equilibrium unit-cell volume 𝑉, atomic displacements in percent of 𝑏, and total-energy
difference Δ𝐸 per formula unit with respect to cubic KNbO3 calculated with different exchange-correlation functionals for the orthorhombic
phase of KNbO3 . Data in brackets represent relative deviations from the experimental values in percent.

𝐴𝑚𝑚2 LDA PBE PBEsol AM05 RTPSS Expt.a

𝑎 (Å) 3.940 (−0.83) 3.983 (+0.25) 3.961 (−0.30) 3.958 (−0.38) 3.984 (+0.28) 3.973
𝑏 (Å) 3.971 (−1.59) 4.100 (+1.61) 4.021 (−0.35) 4.043 (+0.20) 4.048 (+0.32) 4.035
𝛼 (∘ ) 89.91 (+0.20) 89.66 (−0.08) 89.82 (+0.10) 89.76 (+0.03) 89.81 (+0.09) 89.73
𝑎/𝑏 0.9922 (+0.77) 0.9715 (−1.33) 0.9851 (+0.05) 0.9790 (−0.57) 0.9842 (−0.04) 0.9846
3
𝑉 (Å ) 62.13 (−3.94) 66.95 (+3.51) 64.04 (−0.99) 64.70 (+0.03) 65.28 (+0.93) 64.68
ΔK 0.632 1.095 0.877 0.987 0.971 0.974
ΔNb 0 0 0 0 0 0
ΔO1 1.793 2.942 2.306 2.618 2.381 2.568
ΔO2 1.615 2.768 2.131 2.432 2.216 2.413
ΔO3 1.615 2.768 2.131 2.432 2.216 2.413
Δ𝐸 (eV) −0.0080 −0.0617 −0.0245 −0.0401 −0.0289
a
Ref. [41]; measured at 295 K.

is underlined by improved 𝑎/𝑏 ratios with PBEsol (0.9851) along 𝑏 and 𝑐 directions and no detectable component along
and RTPSS (0.9842) which show virtually identical but 𝑎. Compared to PBEsol and RTPSS, AM05 shows a slight
opposing minimal deviations from the experimental value tendency towards larger atomic displacements. Although
(0.9846). While all functionals predict a very similar small there are no directly comparable experimental values, the
reduction of angle 𝛼, the experimental value of 89.73∘ is most calculated displacements are compatible with those of Zhang
accurately reproduced by AM05 (89.76∘ ). AM05 also yields et al. [48] based on density-functional theory in combination
the equilibrium volume (64.70) closest to experiment (64.68). with the otherwise rarely used Wu-Cohen functional [49].
The three functionals PBEsol, AM05, and RTPSS not only Finally, for the rhombohedral phase, we set the starting
yield results in reasonable quantitative agreement with the values for the displacements of all atoms except Nb to two
experimental data but also give a good account of the atomic percent along each lattice vector, because the structure tends
displacements, although the best description of the internal to relax to the cubic phase for too small initial displacements.
degrees of freedom is provided by the AM05 parametrization. The resulting optimized values are listed in Table 5. The lattice
Overall, AM05 provides the best balanced description of all parameter 𝑎 and angle 𝛼 obtained from PBEsol, AM05, and
structural degrees of freedom in orthorhombic KNbO3 . RTPSS are overall in good agreement with the experimental
The results for the newly discovered monoclinic phase data. The performance of LDA and PBE follows that for the
depend sensitively on the starting parameters for the other phases and shows characteristic underestimation and
atomic displacements, as the structure tends to relax to the overestimation of the lattice parameters, respectively. Our
orthorhombic phase for poorly chosen initial displacements. calculated atomic displacements are notably larger than those
The only previous theoretical study of KNbO3 in the mon- reported by Fontana et al. in 1984 [41], especially for the oxy-
oclinic phase was reported by Zhang et al. [48]. In our gen atoms. The reason for this discrepancy remains unclear.
simulations, we take the displacements given in [48] as the On one hand, the theoretical results are very consistent for
initial configuration. In this way, we eventually obtain a all functionals considered here and thus appear robust. On
properly relaxed monoclinic structure, which corresponds to the other hand, the low-temperature rhombohedral phase
a local minimum of the total-energy landscape, indicating has received relatively little attention in the literature so
that the monoclinic phase is metastable not only for KNbO3 far, so that there are no other, more recent experimental
nanomaterials but also for the bulk. The results listed in characterizations of this structure which could be used for
Table 4 show that the performance of the different functionals further comparison. In our simulations, all displacements are
is the same as that for the related orthorhombic phase, which predominantly along the spatial diagonal of the unit cell with
occurs in the same temperature range. The universal trends roughly equal components along all three lattice vectors.
of underestimation and overestimation of lattice parameters
by the LDA and PBE functionals are clearly reflected in equi- 3.2. Phase Stability. For the purpose of comparison, Table 6
3 3
librium volumes of 62.13 Å and 66.99 Å which fall below summarizes our results for the total-energy differences Δ𝐸
3
or exceed the experimental volume of 64.86 Å , respectively. per formula unit for the various phases of KNbO3 , which
Surprisingly, LDA performs best for angle 𝛼 as well as for correspond to the formation energies relative to the cubic
the ratios 𝑎/𝑏 and 𝑎/𝑐. A notable improvement over the LDA polymorph. Although the numerical differences are gen-
and PBE values is observable for the other functionals, but erally small, we find that the rhombohedral structure is
no clear trend emerges. PBEsol, AM05, and RTPSS provide energetically favored by all exchange-correlation functionals
virtually the same accuracy for the lattice parameters. The considered in this work. As the calculations refer to 0 K, this
atomic displacements have again roughly equal components is gratifyingly consistent with the experimental fact that the
6 Advances in Materials Science and Engineering

Table 4: Lattice parameters 𝑎, 𝑏, and 𝑐, angle 𝛼, ratios 𝑎/𝑏 and 𝑎/𝑐, equilibrium unit-cell volume 𝑉, atomic displacements in percent of 𝑏, and
total-energy difference Δ𝐸 per formula unit with respect to cubic KNbO3 calculated with different exchange-correlation functionals for the
monoclinic phase of KNbO3 . Data in brackets represent relative deviations from the experimental values in percent.

𝑃𝑚 LDA PBE PBEsol AM05 RTPSS Expt.a

𝑎 (Å) 3.940 (−1.28) 3.983 (−0.20) 3.961 (−0.75) 3.958 (−0.83) 3.984 (−0.18) 3.991
𝑏 (Å) 3.972 (−1.83) 4.104 (+1.43) 4.023 (−0.57) 4.045 (−0.02) 4.049 (+0.07) 4.046
𝑐 (Å) 3.970 (−1.17) 4.098 (+2.02) 4.020 (+0.07) 4.041 (+0.60) 4.047 (+0.75) 4.017
𝛼 (∘ ) 89.91 (+0.02) 89.66 (−0.26) 89.82 (−0.08) 89.76 (−0.14) 89.81 (−0.09) 89.89
𝑎/𝑏 0.9919 (+0.56) 0.9705 (−1.61) 0.9846 (−0.18) 0.9785 (−0.80) 0.9839 (−0.25) 0.9864
𝑎/𝑐 0.9924 (−0.11) 0.9719 (−2.17) 0.9853 (−0.83) 0.9795 (−1.41) 0.9844 (−0.92) 0.9935
3
𝑉 (Å ) 62.13 (−4.21) 66.99 (+3.28) 64.06 (−1.23) 64.70 (−0.25) 65.28 (+0.65) 64.86
ΔK 1.260 2.194 1.736 1.971 1.942
ΔNb 0 0 0 0 0
ΔO1 3.571 5.906 4.626 5.234 4.771
ΔO2 3.208 5.583 4.262 4.876 4.443
ΔO3 3.197 5.679 4.263 4.890 4.446
Δ𝐸 (eV) −0.0085 −0.0623 −0.0250 −0.0405 −0.0294
a
Ref. [47]; measured at 298 K.

Table 5: Lattice parameter 𝑎, angle 𝛼, equilibrium unit-cell volume 𝑉, atomic displacements in percent of 𝑎, and total-energy difference
Δ𝐸 per formula unit with respect to cubic KNbO3 calculated with different exchange-correlation functionals for the rhombohedral phase of
KNbO3 . Data in brackets represent relative deviations from the experimental values in percent.

R3m LDA PBE PBEsol AM05 RTPSS Expt.a

𝑎 (Å) 3.961 (−1.37) 4.059 (+1.07) 4.001 (−0.37) 4.014 (−0.05) 4.027 (+0.27) 4.016
𝛼 (∘ ) 89.95 (+0.03) 89.80 (−0.13) 89.90 (−0.02) 89.86 (−0.07) 89.95 (+0.03) 89.92
3
𝑉 (Å ) 62.15 (−4.05) 66.87 (+3.24) 64.05 (−1.11) 64.67 (−0.15) 65.30 (+0.82) 64.77
ΔK 1.296 2.281 1.794 2.046 2.037 1.300
ΔNb 0 0 0 0 0 0
ΔO1 3.340 5.548 4.360 4.945 4.580 0.301
ΔO2 3.340 5.548 4.360 4.945 4.580 0.333
ΔO3 3.340 5.548 4.360 4.945 4.580 0.333
Δ𝐸 (eV) −0.0087 −0.0623 −0.0256 −0.0410 −0.0305
a
Ref. [41]; measured at 230 K.

Table 6: Calculated total-energy differences Δ𝐸 with respect to the full incorporation of phonon contributions [54] which
the cubic phase in eV per formula unit with different exchange- goes beyond the present study, however.
correlation functionals for the five phases of KNbO3 . Figure 3 illustrates the correlation between the unit-cell
volume and the relative formation energies. The remarkably
Phase LDA PBE PBEsol AM05 RTPSS
narrow spread of the LDA results and the much larger
𝑃𝑚3𝑚 0 0 0 0 0
spread of the PBE results for the different phases are clearly
P4mm −0.0069 −0.0522 −0.0201 −0.0330 −0.0231 visible. This is in accordance with our earlier observation
Amm2 −0.0080 −0.0617 −0.0245 −0.0401 −0.0289 that the LDA tends to underestimate and PBE tends to
𝑃𝑚 −0.0085 −0.0623 −0.0250 −0.0405 −0.0294 overestimate symmetry-breaking deformations relative to the
R3m −0.0087 −0.0623 −0.0256 −0.0410 −0.0305 cubic structure. The PBEsol and AM05 results lie between the
LDA and PBE values, indicating a more balanced description
of the structure and energetics. The RTPSS results span a
rhombohedral phase is stable in the low-temperature limit. similar energy range as PBEsol and AM05 but are shifted
It is followed, in ascending order, by the monoclinic, the to slightly larger volumes. In addition, Figure 3 shows an
orthorhombic, the tetragonal, and finally the cubic structure, almost linear dependence of the formation energies on the
consistent among all functionals. The energetic ordering of unit-cell volume, suggesting a dominant volume effect on the
the different phases thus follows the sequence in which they phase stability of the KNbO3 polymorphs. Similar trends in
become stable for increasing temperature. A proper ab initio the correlation between formation energies and equilibrium
description of the phase transitions requires a numerical volume have also been observed for other oxide materials that
evaluation of the temperature-dependent free energies with exhibit extended structure polymorphism [55, 56].
Advances in Materials Science and Engineering 7

0.00 Table 7: Complete list of calculated PBEsol lattice parameters for

the five different phases of KNbO3 , which are used in the HSE
−0.01 electronic-structure calculations.
−0.02 Phase 𝑃𝑚3𝑚 P4mm Amm2 𝑃𝑚 R3m
𝑎 (Å) 3.985 4.058 3.961 3.961 4.001
−0.03
ΔE (eV)

𝑏 (Å) 3.985 3.969 4.021 4.023 4.001

−0.04 𝑐 (Å) 3.985 3.969 4.021 4.020 4.001
𝛼 (∘ ) 90 90 89.82 89.82 89.90
−0.05
𝛽 (∘ ) 90 90 90 90 89.90
−0.06 𝛾 (∘ ) 90 90 90 90 89.90
−0.07

62 63 64 65 66 67 68
3 this does not disqualify any of the considered exchange-
Volume (Å )
correlation functionals from applications to KNbO3 . This
LDA AM05 happenstance of correct phase ordering independent of
PBE RTPSS specific functionals, especially among structurally complex
PBEsol oxides characterized by polymorph-rich phase diagrams, is
Figure 3: Correlation between the equilibrium unit-cell volume 𝑉
in fact rather fortuitous. In the case of TiO2 , for example,
and the total-energy difference per formula unit Δ𝐸 with respect to common local (LDA) and semilocal (PBE and PBEsol) as
the cubic phase calculated with different exchange-correlation func- well as hybrid functionals fail to reproduce the experi-
tionals. The solid lines are intended to guide the eyes. The dashed mentally observed energy ordering among the three main
bars indicate the total volume and energy ranges encompassing all crystallographic phases rutile, anatase, and brookite, which
five KNbO3 phases for a particular functional. structurally differ solely by distortions and rearrangements of
the TiO6 octahedral building blocks [60–63]. Furthermore,
the choice of exchange-correlation functionals in general,
as well as the fraction of exact exchange in various hybrid
3.3. Functional Performance. The underestimation and over- functionals, has been found to alter the phase stability of
estimation of lattice parameters by LDA and by conventional oxide systems [63, 64].
generalized gradient approximations like PBE, respectively, The body of data for the atomic structure and phase
observed consistently throughout this work for all investi- energetics suggests that there is no good justification to
gated KNbO3 polymorphs, are well known and documented stick to the LDA or to conventional GGAs when dealing
for a large variety of materials [45, 57, 58]. In particular, with complex oxide materials. More recent parametrizations
these trends were confirmed in a DFT study of various like PBEsol or AM05, which specifically aim at an accurate
ferroelectric perovskites by Tinte et al. [59]. Nevertheless, description of solid-state systems, as well as meta-GGAs like
the reasonable overall agreement with experimental data RTPSS, constructed to satisfy a maximum of exact constraints
together with the well-understood performance in numerical for the exchange-correlation functional without empirical
simulations makes these established functionals a sound ref- parameters, offer a reliable and numerically efficient way to
erence for the benchmarking of other exchange-correlation improve the accuracy of state-of-the-art ab initio simulations.
functionals. The improved prediction of structural data by All three functionals provide a sound basis for subsequent
PBEsol, AM05, and RTPSS observed here for KNbO3 is calculations. In the following, we largely restrict the analysis
likewise in accordance with recent studies for simple metallic of the electronic structure to the PBEsol geometries summa-
and semiconducting systems [36, 42–46, 58], but our results rized in Table 7, as we do not expect qualitatively different
demonstrate that this extends to oxides with various stable trends for geometries from AM05 or RTPSS. In view of future
and metastable polymorphs of complex symmetry, which studies, these functionals seem to be well suited as a starting
exhibit manifold internal degrees of freedom. We note in point not only for hybrid-functional DFT but also for higher-
this context that it is also possible to improve conventional level perturbative approaches within the framework of many-
semilocal functionals through restriction to specific classes of body perturbation theory [65]. The analysis of vibrational
materials by introducing system-dependent parameters that properties in KNbO3 will equally benefit from the qualitative
modify the localization of the exchange-correlation hole [59], and quantitative reliability provided by these functionals, as
an idea that is not strictly ab initio but presently has a strong recently demonstrated for the related material LiNbO3 [24].
influence on the construction and use of hybrid functionals.
The energetic ordering (Table 6) of the five KNbO3 3.4. Band Structures. We now proceed to analyzing the effects
polymorphs is identical for all considered functionals and, of the structural deformations in the various phases on the
most importantly, in accordance with the observed experi- electronic band structures. In a preliminary step, we study
mental sequence. Hence, even though the LDA energy differ- the influence of the structural parameters and the exchange-
ences are close to the overall numerical accuracy of current correlation functionals for the cubic phase, which serves as
implementations of plane-wave DFT for solid-state systems, a test case. Table 8 shows results for the fundamental band
8 Advances in Materials Science and Engineering

4.5 Table 8: Fundamental 𝑅V → Γ𝑐 band gaps for the cubic phase of

4.0 KNbO3 calculated with different lattice parameters and exchange-
correlation functionals (see Figure 4).
3.5
Band gap (eV)

3.0 Geometry Functional Band gap (eV)

2.5 Expt. PBE 1.426
2.0 LDA LDA 1.377
1.5 PBE 1.426
PBE
1.0 PBEsol 1.125
25 30 35 40 45 50 PBE 1.540
PBEsol
Fraction of exact exchange (%) PBEsol 1.396
HSE-x PBE HSE-25 2.701
LDA PBEsol HSE-29 3.080
Figure 4: Fundamental 𝑅V → Γ𝑐 band gaps for the cubic phase of HSE-30 3.138
KNbO3 calculated with different exchange-correlation functionals PBEsol HSE-31 3.194
(see Table 8). The HSE-𝑥 band gaps are calculated for PBEsol equi- HSE-32 3.252
librium geometries and depend on the fraction of exact exchange 𝑥. HSE-33 3.346
The solid lines are intended to guide the eyes. The dashed lines mark
HSE-50 4.176
the size of the fundamental LDA, PBE, and PBEsol band gaps.

Table 9: Fundamental band gaps for the different phases of KNbO3

with the HSE-30 functional from the band structures in Figure 5
gaps with different combinations of structural parameters compared to experimental values.
(see Table 1) and functionals. In all cases, the band gap
is predicted to be indirect and corresponds to a transition HSE-30 Expt.
Phase
between the valence-band maximum at 𝑅, the corner of the Transition Band gap (eV) Band gap (eV)
cubic Brillouin zone, and the conduction-band minimum at 𝑃𝑚3𝑚 𝑅V → Γ𝑐 3.14 3.14a , 3.16b , 3.24c
Γ, its center. Furthermore, we find that the band gap is not P4mm 𝑀V → Γ𝑐 3.23 3.08c , 3.30d
very sensitive to the lattice parameter and that, as expected, all Amm2 𝑇V → Γ𝑐 3.59 3.15c , 3.17e , 3.25f
(semi)local functionals like LDA, PBE, or PBEsol systemati- Pm 𝐴 V → Γ𝑐 3.63 3.09e , 3.15f
cally underestimate the reported experimental values, which R3m 𝑍V → Γ𝑐 3.80
lie between 3.14 eV [50] and 3.24 eV [52]. According to the a
Ref. [50]; optical gap from absorption spectrum.
preceding section, we consistently use structural parameters b
Ref. [51]; optical gap from UV-vis diffuse reflectance spectrum.
from PBEsol (see Table 7) from now on, and we perform c
Ref. [52]; optical gap from absorption spectrum.
a series of additional band-structure calculations with the d
Ref. [53]; optical gap from X-ray photoelectron spectrum.
e
nonlocal HSE functional. In our notation, HSE-𝑥 means that Ref. [47]; optical gap from absorption spectrum.
the HSE functional is used with 𝑥 percent exact exchange, f
Ref. [48]; optical gap from diffuse reflectance UV-vis spectroscopy.
which we treat as a tunable parameter. Depending on 𝑥, the
hybrid-functional band gap can be positioned between the
lower limit of (semi)local DFT and the upper limit of Hartree- other phases, it represents a reasonable approach to study
Fock theory. In Figure 4, the dependence of the fundamental physical trends among the various polymorphs.
band gap of cubic KNbO3 on the fraction of exact exchange Figure 5 shows the calculated band structures for the
is illustrated for 25 ≤ 𝑥 ≤ 50. In the most relevant region five different phases of KNbO3 . Furthermore, in Table 9,
28 ≤ 𝑥 ≤ 32, close to the experimentally reported band we compare the fundamental band gaps with experimental
gap, an almost linear dependence on 𝑥 is observed. The LDA, values. The calculated HSE-30 band gaps show a broad spread
PBE, and PBEsol band gaps are shown for comparison and of 0.66 eV between the fundamental band gaps of cubic
exhibit an underestimation by a factor of one half. According (3.14 eV) and rhombohedral (3.80 eV) KNbO3 . It should be
to Table 8, a choice of 30% exact exchange leads to much noted that our results refer to electronic band gaps, which can
better agreement with the experimental band gap than the be measured, in principle, by photoemission spectroscopy.
standard setting of 25%, so that we use the HSE-30 variant No such measurements have been reported for KNbO3 so
in the following. The improved description of band gaps and far, however. Instead, the experimental values listed in the
the electronic structure in general with an increased fraction table are deduced from optical spectroscopies and thus refer
of exact exchange in HSE-type hybrid functionals has previ- to optical band gaps. The difference between the two equals
ously been demonstrated for oxide semiconductors [61] and the exciton binding energy, whose magnitude is unknown
nitride semiconductor alloys [66, 67], for instance. Wang et al. for KNbO3 , but detailed investigations of the related material
[68] also noted that an increased fraction of exact exchange LiNbO3 [25] suggest that it may be significant. In addition,
improves the band gap of KNbO3 . Although adjusting the the experimental values include thermal effects due to ther-
fraction of exact exchange for a single polymorph might not mal expansion and electron-phonon coupling, particularly in
yield perfect agreement between experiment and theory for the high-temperature phases, as well as size and geometry
Advances in Materials Science and Engineering 9

kz
10 10

Energy (eV)

Energy (eV)
R 5 5

Γ 0 0
X ky
kx M
−5 −5

Γ X M Γ R X DOS
(a.u.)

(a)

kz 10 10
R

Energy (eV)

Energy (eV)
5 5

Γ X 0 0
ky
M
kx −5 −5

Γ X M Γ R X DOS
(a.u.)

(b)

kz 10 10
Energy (eV)

Energy (eV)
Z 5 5
T R
0 0
Γ
Y S ky −5 −5
kx
S R Z Γ Y T DOS
(a.u.)

(c)

kz 10 10
Z
C
Energy (eV)

Energy (eV)
5 5

A Γ 0 0
kx B Y ky
−5 −5

A B Γ Y C Z DOS
(a.u.)

(d)

kz 10 10
Z
Energy (eV)

Energy (eV)

F Λ 5 5
L
0 0
Γ ky
kx
−5 −5

Z F L Γ Λ Z DOS
(a.u.)
Total Op
Nb d
(e)

Figure 5: Band structures calculated with the HSE-30 functional for the cubic (a), tetragonal (b), orthorhombic (c), monoclinic (d), and
rhombohedral (e) phases of KNbO3 with corresponding densities of states (DOS) in arbitrary units. The grey-shaded areas indicate the PBE
band structures for comparison, and the dashed horizontal lines mark the valence-band maxima and conduction-band minima. All band-
structure calculations are based on structural data from PBEsol (see Table 7). On the left-hand side, the chosen paths within the first Brillouin
zone are indicated by thick dash-dotted lines.
10 Advances in Materials Science and Engineering

Table 10: Calculated valence-band edge 𝐸V (k) and conduction-band edge 𝐸c (k) at high-symmetry k points for the different phases of KNbO3 .
Energies are given in eV relative to the position of the valence-band maximum. The valence-band maximum and the conduction-band
minimum are italicized.
Cubic Tetragonal Orthorhombic Monoclinic Rhombohedral
(𝑃𝑚3𝑚) (P4mm) (Amm2) (𝑃𝑚) (R3m)
k 𝐸V (k) 𝐸𝑐 (k) k 𝐸V (k) 𝐸𝑐 (k) k 𝐸V (k) 𝐸𝑐 (k) k 𝐸V (k) 𝐸𝑐 (k) k 𝐸V (k) 𝐸𝑐 (k)
Γ −0.72 3.14 Γ −0.90 3.23 𝑆 −1.23 4.10 𝐴 0 6.34 𝑍 0 7.35
𝑋 −0.94 3.27 𝑋 −1.24 3.96 𝑅 −0.04 6.30 𝐵 −0.97 3.92 𝐹 −0.04 6.13
𝑀 −0.04 6.11 𝑀 0 6.26 𝑍 −1.01 3.87 Γ −0.88 3.63 𝐿 −1.04 3.98
𝑅 0 7.51 𝑅 −0.21 6.16 Γ −0.92 3.59 𝑌 −1.19 4.14 Γ −0.97 3.80
𝑌 −0.04 6.30 𝐶 −0.004 6.34 Λ −0.60 5.71
𝑇 0 7.36 𝑍 −0.97 3.90

effects, because the experiments were typically not carried out monoclinic phase, based on density-functional theory. As the
for bulk samples but for nanostructures, such as microcubes energetic separation and the structural differences between
in [52] or nanowires in [48]. Nanoparticles of TiO2 , for the phases are small and thus potentially sensitive to details
instance, exhibit much smaller band gaps than in the bulk of the computational procedure, we have placed special
[69]. For all these reasons, the comparison in Table 9 can emphasis on the influence of different approximations for the
only be indicative. No experimental band gap is available for exchange-correlation functional. To this end, we compared
the rhombohedral phase, but our results are consistent with the LDA with a range of GGAs, a meta-GGA, and, in the
earlier calculations by Wang and Rappe [70] which yielded a case of band structures, a HSE-type hybrid functional. As
fundamental band gap of 3.33 eV with the HSE-25 functional an important result, we find that all functionals predict the
and LDA structural data; our own HSE-25 calculation with same energetic ordering of the phases, which, moreover,
structural data from PBEsol for the rhombohedral phase coincides with the order in which the phases become stable
yields 3.38 eV. for increasing temperature. Our results further confirm that
The graphical band structures in Figure 5 as well as the the monoclinic configuration, which has so far only been
numerical data in Table 10, which lists energy values of the observed in synthesized nanostructures, is also metastable
highest valence band and the lowest conduction band at in the bulk. Among the considered functionals, PBEsol as
selected high-symmetry k points, confirm that in all phases well as AM05 and RTPSS yield structural data in consistently
there are states elsewhere in the Brillouin zone which are good quantitative agreement with experimental measure-
nearly degenerate with the valence-band maximum or the ments for all phases; in contrast, LDA and PBE systemati-
conduction-band minimum. For example, in the orthorhom- cally underestimate and overestimate the lattice parameters,
bic phase that is stable at room temperature, the valence-band respectively, and are also less accurate with respect to the
edge at 𝑌 lies only 0.04 eV below the valence-band maximum, atomic displacements inside the unit cell. For this reason,
which is located at 𝑇. It is expected that such near degen- we even-handedly recommend PBEsol, AM05, or RTPSS
eracies strongly influence the optical properties close to the as reasonable starting points for ab initio investigations
absorption edge due to strong mixing of electronic transitions with higher-level perturbative methods that require accurate
at different k points in the optical response. The correspond- equilibrium geometries. In particular, phonons, which not
ing densities of states (DOS) in Figure 5 share many similar only are crucial in the context of phase transitions but also
features, independent of the particular polymorph. The DOS play a significant role for the temperature dependence of the
contributions from the K atoms are negligible around the electronic and optical properties due to the electron-phonon
band edges in all phases. The topmost valence bands are coupling, are sensitive to small changes in the atomic spacing.
consistently formed by oxygen 2𝑝 states and give rise to two Our band-structure calculations performed with the HSE
prominent peaks close to the valence-band edge: the first hybrid functional and 30% exact exchange confirm that
peak arises from the valence-band maximum and the nearly the fundamental band gap is indirect for all phases. The
degenerate electronic states at other high-symmetry k points, variation in the size of the gap amounts to 0.66 eV through
while the second peak originates from an almost degenerate the successive phase transitions. In particular, in the room-
continuum of states around Γ; its energetic distance from the temperature orthorhombic phase, the band gap is appreciably
conduction-band edge equals the direct band gap at Γ for all larger than in the tetragonal phase, which is often used as
KNbO3 phases. The lowest conduction bands, in contrast, are a structurally simpler substitute in theoretical simulations.
mainly formed by niobium 4𝑑 states. For all phases, our results further indicate a range of near
degeneracies of band-edge states in different parts of the
4. Conclusions Brillouin zone which are expected to be relevant for the linear
and nonlinear optical properties of KNbO3 . More elaborate
In summary, we have performed a comprehensive analysis of ab initio band-structure calculations with a proper treatment
the structural and electronic properties of KNbO3 in its five of quasiparticle and thermal effects, which go beyond the
known phases, including the only very recently discovered present study, are highly desirable, however.
Advances in Materials Science and Engineering 11

Competing Interests [14] T. Neumann, G. Borstel, C. Scharfschwerdt, and M. Neumann,

“Electronic structure of KNbO3 and KTaO3 ,” Physical Review B,
The authors declare that there are no competing interests vol. 46, no. 17, pp. 10623–10628, 1992.
regarding the publication of this paper. [15] S. Cabuk, “Electronic structure and optical properties of
KNbO3 : first principles study,” Optoelectronics and Advanced
Acknowledgments Materials—Rapid Communications, vol. 1, no. 3, pp. 100–107,
2007.
The authors gratefully acknowledge financial support from [16] A. V. Postnikov, T. Neumann, G. Borstel, and M. Methfessel,
the Deutsche Forschungsgemeinschaft (DFG) via Sonder- “Ferroelectric structure of KNbO3 and KTaO3 from first-
forschungsbereich TRR 142. principles calculations,” Physical Review B, vol. 48, no. 9, pp.
5910–5918, 1993.
[17] R. I. Eglitis, A. V. Postnikov, and G. Borstel, “Semiempirical
References Hartree-Fock calculations for KNbO3 ,” Physical Review B, vol.
[1] C.-M. Fang and R. Ahuja, “Structures and stability of ABO3 54, no. 4, pp. 2421–2427, 1996.
orthorhombic perovskites at the Earth’s mantle conditions [18] E. A. Kotomin, R. I. Eglitis, A. V. Postnikov, G. Borstel, and N. E.
from first-principles theory,” Physics of the Earth and Planetary Christensen, “First-principles and semiempirical calculations
Interiors, vol. 157, no. 1-2, pp. 1–7, 2006. for bound-hole polarons in KNbO3 ,” Physical Review B, vol. 60,
[2] R. H. Mitchell, Perovskites: Modern and Ancient, Almaz Press, no. 1, pp. 1–5, 1999.
Thunder Bay, Canada, 2002. [19] N. E. Christensen, E. A. Kotomin, R. I. Eglitis et al., “Quantum
[3] R. J. D. Tilley, Perovskites: Structure-Property Relationships, John mechanical modelling of pure and defective KNbO3 per-
Wiley & Sons, Chichester, UK, 2016. ovskites,” in Defects and Surface-Induced Effects in Advanced
[4] B. T. Matthias, “New ferroelectric crystals,” Physical Review, vol. Perovskites, G. Borstel, A. Krumins, and D. K. Millers, Eds., pp.
75, no. 11, p. 1771, 1949. 3–16, Kluwer, Dordrecht, The Netherlands, 2000.
[5] S. Sanna, C. Thierfelder, S. Wippermann, T. P. Sinha, and W. G. [20] P. Hohenberg and W. Kohn, “Inhomogeneous electron gas,”
Schmidt, “Barium titanate ground- and excited-state properties Physical Review, vol. 136, no. 3B, pp. B864–B871, 1964.
from first-principles calculations,” Physical Review B, vol. 83, no. [21] W. Kohn and L. J. Sham, “Self-consistent equations including
5, Article ID 054112, 2011. exchange and correlation effects,” Physical Review, vol. 140, no.
[6] A. W. Hewat, “Cubic-tetragonal-orthorhombic-rhombohedral 4A, pp. A1133–A1138, 1965.
ferroelectric transitions in perovskite potassium niobate: neu- [22] C. Thierfelder, S. Sanna, A. Schindlmayr, and W. G. Schmidt,
tron powder profile refinement of the structures,” Journal of “Do we know the band gap of lithium niobate?” Physica Status
Physics C: Solid State Physics, vol. 6, no. 16, pp. 2559–2572, 1973. Solidi (C) Current Topics in Solid State Physics, vol. 7, no. 2, pp.
[7] S. Kim, J.-H. Lee, J. Lee et al., “Synthesis of monoclinic potas- 362–365, 2010.
sium niobate nanowires that are stable at room temperature,” [23] A. Riefer, S. Sanna, A. Schindlmayr, and W. G. Schmidt, “Optical
Journal of the American Chemical Society, vol. 135, no. 1, pp. 6– response of stoichiometric and congruent lithium niobate from
9, 2013. first-principles calculations,” Physical Review B, vol. 87, no. 19,
[8] S. Simsek, H. Koc, V. A. Trepakov, A. M. Mamedov, and E. Article ID 195208, 2013.
Ozbay, “Electron spectroscopy and the electronic structure of [24] M. Friedrich, A. Riefer, S. Sanna, W. G. Schmidt, and A. Schindl-
KNbO3 : first principle calculations,” Ferroelectrics, vol. 461, no. mayr, “Phonon dispersion and zero-point renormalization of
1, pp. 99–105, 2014. LiNbO3 from density-functional perturbation theory,” Journal
[9] L. Grigorjeva, D. K. Millers, V. Pankratov et al., “Experimental of Physics Condensed Matter, vol. 27, no. 38, Article ID 385402,
and theoretical studies of polaron optical properties in KNbO3 2015.
perovskite,” Solid State Communications, vol. 129, no. 11, pp. 691– [25] A. Riefer, M. Friedrich, S. Sanna, U. Gerstmann, A. Schindl-
696, 2004. mayr, and W. G. Schmidt, “LiNbO3 electronic structure: many-
[10] E. A. Kotomin, R. I. Eglitis, G. Borstel, L. Grigorjeva, D. Millers, body interactions, spin-orbit coupling, and thermal effects,”
and V. Pankratov, “Theoretical and experimental study of pri- Physical Review B, vol. 93, no. 7, Article ID 075205, 2016.
mary radiation defects in KNbO3 perovskite crystals,” Nuclear
[26] L. J. Sham and M. Schlüter, “Density-functional theory of the
Instruments and Methods in Physics Research B, vol. 166-167, pp.
energy gap,” Physical Review Letters, vol. 51, no. 20, pp. 1888–
299–304, 2000.
1891, 1983.
[11] E. E. Krasovskii, O. V. Krasovska, and W. Schattke, “Ab initio
[27] J. P. Perdew, “Density functional theory and the band gap prob-
calculation of the optical and photoelectron spectra of KNbO3
lem,” International Journal of Quantum Chemistry, vol. 28, no.
and KTaO3 ,” Journal of Electron Spectroscopy and Related
19S, pp. 497–523, 1985.
Phenomena, vol. 83, no. 2-3, pp. 121–127, 1997.
[12] S. Torbrügge, M. Imlau, B. Schoke et al., “Optically generated [28] G. Kresse and J. Furthmüller, “Efficiency of ab-initio total
small electron and hole polarons in nominally undoped and energy calculations for metals and semiconductors using a
Fe-doped KNbO3 investigated by transient absorption spec- plane-wave basis set,” Computational Materials Science, vol. 6,
troscopy,” Physical Review B, vol. 78, no. 12, Article ID 125112, no. 1, pp. 15–50, 1996.
2008. [29] G. Kresse and D. Joubert, “From ultrasoft pseudopotentials to
[13] H. Donnerberg and M. Exner, “Derivation and application of ab the projector augmented-wave method,” Physical Review B, vol.
initio Nb5+ –O2− short-range effective pair potentials in shell- 59, no. 3, pp. 1758–1775, 1999.
model simulations of KNbO3 and KTaO3 ,” Physical Review B, [30] P. E. Blöchl, “Projector augmented-wave method,” Physical
vol. 49, no. 6, pp. 3746–3754, 1994. Review B, vol. 50, no. 24, pp. 17953–17979, 1994.
12 Advances in Materials Science and Engineering

[31] J. P. Perdew and A. Zunger, “Self-interaction correction to [48] T. Zhang, W. Lei, P. Liu et al., “Insights into the structure-photo-
density-functional approximations for many-electron systems,” reactivity relationships in well-defined perovskite ferroelectric
Physical Review B, vol. 23, no. 10, pp. 5048–5079, 1981. KNbO3 nanowires,” Chemical Science, vol. 6, no. 7, pp. 4118–
[32] J. P. Perdew, K. Burke, and Y. Wang, “Generalized gradient 4123, 2015.
approximation for the exchange-correlation hole of a many- [49] Z. Wu and R. E. Cohen, “More accurate generalized gradient
electron system,” Physical Review B, vol. 54, no. 23, pp. 16533– approximation for solids,” Physical Review B, vol. 73, no. 23,
16539, 1996. Article ID 235116, 2006.
[33] J. P. Perdew, A. Ruzsinszky, G. I. Csonka et al., “Restoring the [50] J. W. Liu, G. Chen, Z. H. Li, and Z. G. Zhang, “Hydrothermal
density-gradient expansion for exchange in solids and surfaces,” synthesis and photocatalytic properties of ATaO3 and ANbO3
Physical Review Letters, vol. 100, no. 13, Article ID 136406, 2008. (A = Na and K),” International Journal of Hydrogen Energy, vol.
[34] R. Armiento and A. E. Mattsson, “Functional designed to 32, no. 13, pp. 2269–2272, 2007.
include surface effects in self-consistent density functional [51] T. Su, H. Jiang, H. Gong, and Y. Zhai, “An alternative approach
theory,” Physical Review B, vol. 72, no. 8, Article ID 085108, 2005. of solid-state reaction to prepare nanocrystalline KNbO3 ,”
[35] A. E. Mattsson, R. Armiento, J. Paier, G. Kresse, J. M. Wills, and Journal of Materials Science, vol. 45, no. 14, pp. 3778–3783, 2010.
T. R. Mattsson, “The AM05 density functional applied to solids,”
[52] T. Zhang, K. Zhao, J. Yu et al., “Photocatalytic water splitting for
Journal of Chemical Physics, vol. 128, no. 8, Article ID 084714,
hydrogen generation on cubic, orthorhombic, and tetragonal
2008.
KNbO3 microcubes,” Nanoscale, vol. 5, no. 18, pp. 8375–8383,
[36] J. Sun, M. Marsman, G. I. Csonka et al., “Self-consistent meta-
2013.
generalized gradient approximation within the projector-aug-
mented-wave method,” Physical Review B, vol. 84, no. 3, Article [53] Y. Shiozaki, E. Nakamura, and T. Mitsui, Ferroelectrics and
ID 035117, 2011. Related Substances: Oxides Part 1: Perovskite-Type Oxides and
[37] J. Heyd, G. E. Scuseria, and M. Ernzerhof, “Hybrid functionals LiNbO3 Family, New Series III/36A1, Landolt-Börnstein, 2001.
based on a screened Coulomb potential,” Journal of Chemical [54] M. Friedrich, A. Schindlmayr, W. G. Schmidt, and S. Sanna,
Physics, vol. 118, no. 18, pp. 8207–8215, 2003. “LiTaO3 phonon dispersion and ferroelectric transition calcu-
[38] J. Heyd, G. E. Scuseria, and M. Ernzerhof, “Erratum: hybrid lated from first principles,” Physica Status Solidi B, vol. 253, no.
functionals based on a screened Coulomb potential [Journal 4, pp. 683–689, 2016.
of Chemical Physics 118, 8207 (2003)],” Journal of Chemical [55] J. K. Dewhurst and J. E. Lowther, “High-pressure structural
Physics, vol. 124, no. 21, Article ID 219906, 2006. phases of titanium dioxide,” Physical Review B, vol. 54, no. 6,
[39] J. Heyd and G. E. Scuseria, “Efficient hybrid density func- pp. R3673–R3675, 1996.
tional calculations in solids: assessment of the Heyd-Scuseria- [56] A. Beltrán, L. Gracia, and J. Andrés, “Density functional theory
Ernzerhof screened Coulomb hybrid functional,” Journal of study of the brookite surfaces and phase transitions between
Chemical Physics, vol. 121, no. 3, pp. 1187–1192, 2004. natural titania polymorphs,” The Journal of Physical Chemistry
[40] J. Heyd, J. E. Peralta, G. E. Scuseria, and R. L. Martin, B, vol. 110, no. 46, pp. 23417–23423, 2006.
“Energy band gaps and lattice parameters evaluated with the [57] A. van de Walle and G. Ceder, “Correcting overbinding in local-
Heyd-Scuseria-Ernzerhof screened hybrid functional,” Journal density-approximation calculations,” Physical Review B, vol. 59,
of Chemical Physics, vol. 123, no. 17, Article ID 174101, 2005. no. 23, pp. 14992–15001, 1999.
[41] M. D. Fontana, G. Metrat, J. L. Servoin, and F. Gervais, “Infrared [58] P. Haas, F. Tran, and P. Blaha, “Calculation of the lattice constant
spectroscopy in KNbO3 through the successive ferroelectric of solids with semilocal functionals,” Physical Review B, vol. 79,
phase transitions,” Journal of Physics C: Solid State Physics, vol. no. 8, Article ID 085104, 2009.
17, no. 3, pp. 483–514, 1984.
[59] S. Tinte, M. G. Stachiotti, C. O. Rodriguez, D. L. Novikov, and
[42] G. I. Csonka, J. P. Perdew, A. Ruzsinszky et al., “Assessing
N. E. Christensen, “Applications of the generalized gradient
the performance of recent density functionals for bulk solids,”
approximation to ferroelectric perovskites,” Physical Review B,
Physical Review B, vol. 79, no. 15, Article ID 155107, 2009.
vol. 58, no. 18, pp. 11959–11963, 1998.
[43] J. P. Perdew, A. Ruzsinszky, G. I. Csonka, L. A. Constantin, and
J. Sun, “Workhorse semilocal density functional for condensed [60] J. Muscat, V. Swamy, and N. M. Harrison, “First-principles
matter physics and quantum chemistry,” Physical Review Letters, calculations of the phase stability of TiO2 ,” Physical Review B,
vol. 103, no. 2, Article ID 026403, 2009. vol. 65, no. 22, pp. 4112–4126, 2002.
[44] R. Peverati and D. G. Truhlar, “Performance of the M11-L den- [61] M. Landmann, E. Rauls, and W. G. Schmidt, “The electronic
sity functional for bandgaps and lattice constants of unary and structure and optical response of rutile, anatase and brookite
binary semiconductors,” Journal of Chemical Physics, vol. 136, TiO2 ,” Journal of Physics Condensed Matter, vol. 24, no. 19,
no. 13, Article ID 134704, 2012. Article ID 195503, 2012.
[45] L. He, F. Liu, G. Hautier et al., “Accuracy of generalized gradient [62] F. Labat, P. Baranek, C. Domain, C. Minot, and C. Adamo,
approximation functionals for density-functional perturbation “Density functional theory analysis of the structural and elec-
theory calculations,” Physical Review B, vol. 89, no. 6, Article ID tronic properties of TiO2 rutile and anatase polytypes: perfor-
064305, 2014. mances of different exchange-correlation functionals,” Journal
[46] C. Lin, Y. Zhao, and G. Yin, “Calculation of the lattice constant of Chemical Physics, vol. 126, no. 15, Article ID 154703, 2007.
of solids with the use of valence electron structure parameters,” [63] M. Gerosa, C. E. Bottani, L. Caramella, G. Onida, C. Di
Computational Materials Science, vol. 97, pp. 86–93, 2015. Valentin, and G. Pacchioni, “Electronic structure and phase
[47] P. G. Kang, B. K. Yun, S. Shin et al., “Possible origin of stabi- stability of oxide semiconductors: performance of dielectric-
lized monoclinic structure of KNbO3 nanomaterials at room dependent hybrid functional DFT, benchmarked against GW
temperature,” Materials Science and Engineering B, vol. 210, pp. band structure calculations and experiments,” Physical Review
19–23, 2016. B, vol. 91, no. 15, Article ID 155201, 2015.
Advances in Materials Science and Engineering 13

[64] M. T. Curnan and J. R. Kitchin, “Investigating the energetic

ordering of stable and metastable TiO2 polymorphs using
DFT+U and hybrid functionals,” Journal of Physical Chemistry
C, vol. 119, no. 36, pp. 21060–21071, 2015.
[65] A. Schindlmayr, “The GW approximation for the electronic self-
energy,” in Many-Electron Approaches in Physics, Chemistry and
Mathematics, V. Bach and L. Delle Site, Eds., vol. 29 of Math-
ematical Physics Studies, pp. 343–357, Springer International,
Cham, Switzerland, 2014.
[66] M. Landmann, M. Feneberg, M. Röppischer et al., “Transition
energies and direct-indirect band gap crossing in zinc-blende
Al𝑥 Ga1−𝑥 N,” Physical Review B, vol. 87, no. 19, Article ID 195210,
2013.
[67] C. Mietze, M. Landmann, E. Rauls et al., “Band offsets in
cubic GaN/AlN superlattices,” Physical Review B, vol. 83, no. 19,
Article ID 195301, 2011.
[68] G. Wang, H. Chen, Y. Li, A. Kuang, H. Yuan, and G. Wu, “A
hybrid density functional study on the visible light photocata-
lytic activity of (Mo,Cr)–N codoped KNbO3 ,” Physical Chem-
istry Chemical Physics, vol. 17, no. 43, pp. 28743–28753, 2015.
[69] M. M. Khan, S. A. Ansari, D. Pradhan et al., “Band gap engi-
neered TiO2 nanoparticles for visible light induced photoelec-
trochemical and photocatalytic studies,” Journal of Materials
Chemistry A, vol. 2, no. 3, pp. 637–644, 2014.
[70] F. Wang and A. M. Rappe, “First-principles calculation of the
bulk photovoltaic effect in KNbO3 and (K,Ba)(Ni,Nb)O3−𝛿 ,”
Physical Review B, vol. 91, no. 16, Article ID 165124, 2015.
Journal of International Journal of International Journal of Smart Materials Journal of
Nanotechnology
Hindawi Publishing Corporation
Corrosion
Hindawi Publishing Corporation
Polymer Science
Hindawi Publishing Corporation
Research
Hindawi Publishing Corporation
Composites
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of
Metallurgy

BioMed
Research International
Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 Nanomaterials http://www.hindawi.com Volume 2014

Submit your manuscripts at

https://www.hindawi.com

Journal of Journal of
Materials
Hindawi Publishing Corporation
Nanoparticles
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Nanomaterials
Journal of

Advances in The Scientific International Journal of

Materials Science and Engineering
Hindawi Publishing Corporation
Scientifica
Hindawi Publishing Corporation Hindawi Publishing Corporation
World Journal
Hindawi Publishing Corporation
Biomaterials
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of Journal of Journal of Journal of Journal of

Nanoscience
Hindawi Publishing Corporation
Coatings
Hindawi Publishing Corporation
Crystallography
Hindawi Publishing Corporation
Ceramics
Hindawi Publishing Corporation
Textiles
Hindawi Publishing Corporation
http://www.hindawi.com
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Volume 2014