Theme II: Anion control of polar behavior
Theme Leader: Woodward
The traditional approach to the design of functional materials is to consider
a certain fixed anion lattice and then to vary the nature and relative
amounts of the cations. In this research theme we will explore the feasibility
and consequences of substitution on the anion lattice, as an alternative
and underutilized approach to materials development. There are a number
of mixed anionic systems already known that possess interesting properties.
For example superconductivity has been found in copper oxychlorides[1,2],
nitride-chlorides of zirconium and hafnium[3,4],
and oxyfluorides related to the copper oxide superconductors[5,6].
All of these examples adopt layered structures, which are strongly favored
when the two anions have highly contrasting size and/or polarizabilities.
While layered structural motifs are clearly favorable for the occurrence
of superconductivity[7,8,9],
there are relatively few polar materials that adopt layered structures.
Thus our efforts will be initially directed towards compounds where the
two ions are relatively similar in size, in particular oxynitrides, and
to a lesser extent oxyfluorides and nitride fluorides. As the project
progresses computations will be used to see if layered structures are
inherently incompatible with strong polarity, and if not, to suggest which
systems should be synthetically targeted for further study.
Our work will initially focus on oxynitrides, in which some (but
not all) oxide ions are replaced by nitride ions. The importance of this
class of materials is manifested in the recent discovery (by Woodward)
of high permittivity in the perovskite oxynitrides BaTaO2N
(k
~ 5000) and SrTaO2N (k ~ 2500)[10].
The key role of anion substitution is underscored by the fact that the
all-oxide analogues, KTaO3 and NaTaO3, do not exhibit
unusual dielectric properties[11].
However the detailed origin of the large dielectric response is not yet
well understood, and potential applications have yet to be explored. In
this theme we will explain the fundamental chemical interactions driving
the high permittivity, and develop new oxynitride materials with properties
tailored for specific applications.
Our rationale for replacing oxide ions by nitride ions is twofold. First,
the covalency of the metal-anion bond is increased thereby enhancing the
tendency for SOJT distortions. We will use this effect to induce or enhance
SOJT displacements of cations that normally do not show a strong tendency
to undergo out-of-center displacements (i.e. Ta5+, Zr4+,
etc.). Second, the presence of two distinct types of anions affords the
opportunity to create an acentric environment about a metal cation in
a structure where the cation environment would otherwise be a symmetric
octahedron. Here we will use density functional theory to predict off-centering,
since simple intuition based on symmetry considerations has proved inadequate
in the past. For example in BaTaO2N each Ta-centered octahedron
contains on average four oxide ions and two nitride ions Symmetry suggests
that the Ta5+ ion should displace from the center of its coordination
polyhedron when the nitride ions occupy cis-positions in the octahedron,
but not when they adopt a trans-configuration. However density
functional calculations on ordered variants of BaTaO2N indicate
~ 0.10 Å displacements of the Ta cation[12]
in both cis- and trans-configurations, with the trans-
displacement originating from the strong covalency of the Ta-N interaction.
This type of distortion can be rationalized after the fact, but without
the aid of accurate calculational tools it is difficult to predict in
advance. The ATaO2N (A = Ba, Sr) system offers one example
of how polar displacements can be triggered and controlled through manipulations
of the composition and distribution of the anion lattice; in the CDM we
will search computationally and synthetically for other examples.
Progress in the field has been hampered by two factors, both of which
will be overcome in this collaboration. First, oxynitrides are very difficult
to sinter. Preliminary work in Woodward's group (in collaboration with
the Center for Functional Nanomaterials at Brookhaven National Lab) has
shown that thin films of BaTaO2N can be grown, and that the
resulting films have high dielectric constants[13].
Film orientation, the properties and the potential impurity phases are
very sensitive to the substrate, the oxygen partial pressure and the growth
temperature. Extending the study to related materials, for example SrTaO2N
and CaTaO2N, or ordered oxynitrides, such as La3Ta2MgO6N3
and La2TaMgO5N, vastly increases the number of variables.
Therefore, to realize the potential of these intriguing materials, our
experts in film deposition (Salvador and Stemmer). will collaborate closely
with our chemists. A second issue, is understanding the many variables
in structure and composition space that play pivotal roles, and how to
manipulate these variables to design materials with optimal properties.
For example, in BaTaO2N the anion distribution is disordered
and powder diffraction measurements suggest a centrosymmetric, cubic crystal
structure [10,14,15].
This is in contradiction to the DFT calculations on ordered structures[12]
as well as the dielectric properties, and is likely the result of averaging
over differently oriented polar regions. Recent EXAFS measurements give
strong evidence that the Ta ions are displaced and the average crystal
structures do not represent the true local structure[16].
We will use high level calculations to explore the properties of various
anion arrangements. Another question raised by the initial work is the
role of the A-site cation. Measurements on fairly porous ceramic disks
show different temperature dependencies of the dielectric constant (tck)
in BaTaO2N (tck
~ +1138 ppm/K) and SrTaO2N (tck ~ -739 ppm/K),
raising the possibility that solid solutions might show extremely high
permittivity with almost no temperature dependence. Furthermore, CaTaO2N
does not possess a high permittivity (k ~ 30). While
it is well known that dielectric properties are closely linked to various
types of distortions in perovskites the reasons for strikingly different
behavior in these three compounds are not understood at this time. To
develop a better understanding, Spaldin will perform DFT calculations
to determine the types of displacements that occur in ordered oxynitrides,
and the influence of various displacements on the dielectric response.
The correlations between octahedral tilting, SOJT displacements and dielectric
properties will be determined. Her calculations of structural stabilities
will also help determine the range of structure types and counterions
that can be realized with oxynitrides. Finally, with modern DFT tools,
it is possible, to some degree, to study how the energies and properties
of ordered and disordered materials differ. Spaldin's results will allow
us to to screen for new and interesting oxynitride materials, and determine
which materials are worthwhile candidates for large-scale preparation
efforts. The added efficiency obtained by utilizing computations to guide
synthetic efforts should also allow us to broaden our investigations to
other structure types, such as pyrochlore, Ruddlesden-Popper and Aurivillius
phases.
The initial studies of semiconducting oxynitrides are very exciting. This
class of compounds potentially represents a new class of dielectric and
optical materials. Once formed, the perovskite oxynitrides are stable
in air, water, and even strong acids[14,15],
so the potential for applications is very real. However, the synthesis
can be slow and unpredictable. Consequently, the interplay between synthesis
and computations is necessary to increase efficiency. Furthermore, the
extreme difficulty in preparing either single crystal or highly sintered
ceramics makes thin film deposition essential in order to accurately characterize
properties. The skills in synthesis, thin film growth, electrical and
optical characterization and computation that will be assembled in the
CDM represent an ideal mix of skills to effectively attack this challenging,
but potentially very rewarding area of research. Current research activities:
The preparation of oxynitride targets is diffcult and time-consuming, therefore Woodward and Salvador are
exploring the effect of growth conditions on the stability and stoichiometry of oxynitride thin films using PLD
from oxide targets. The work of graduate student Young-Il Kim (Woodward group) has resulted in the growth
of thin film specimens of BaTaO2N perovskites on SrTiO3/SrRuO3 substrates using PLD from a target of
the same composition (in collaboration with Weidong Si and Tom Vogt at BNL). In a collaboration between
Woodward and Salvador, thin films of SrNbO2N have been deposited from an oxide target (Sr2Nb2O7), and are
now being prepared for physical property measurements. The SrNbO2N was found to form as both epitaxial and
polycrystalline films when deposited from the Sr2Nb2O7 target in a N2 atmosphere at 700 ºC. The stability of
polycrystalline films implies that the phase is stable in the ablation conditions and is not an artifact of epitaxial
stabilization; consequently other oxynitrides are likely to be easily synthesized. Ablation targets have therefore
been prepared for other phases, and synthesis conditions of LaTiO2N, LaZrO2N, and LaSnO2N are currently
being optimized. Characterization of the dielectric properties of all the oxynitrides (for Theme III) is in progress. Research findings Pulsed laser deposition from oxide (rather than oxynitrides) targets is an effective way
of growing new oxynitrides. For example SrNbO2N can be formed as both epitaxial and polycrystalline
films when deposited using pulsed laser deposition with a Sr2Nb2O7 ablation target in a nitrogen
atmosphere at ºC. The stability of polycrystalline films implies that the phase is stable in the ablation conditions and is not an
artifact of epitaxial stabilization. This suggests that other oxynitrides are likely to be easily
synthesized.
References