

The low dielectric constant of silicon dioxide has been recognized as one of the fundamental limits on current SiO_2/Si semiconductor technology. Much recent effort on alternate gate dielectrics has been focused on high-\epsilon metal oxides, which might be used to provide physically thicker dielectrics, thus reducing leakage and enhancing gate capacitance. Zirconia (ZrO_2) is a promising candidate because of its good dielectric properties and thermal stability in direct contact with Si. Zirconia has three low-pressure structural phases, namely cubic, tetragonal, and monoclinic. We have studied the dielectric properties of these three polymorphs, including the lattice response as well as the purely electronic response, in the framework of a first-principles ultrasoft-pseudopotential approach.
[E18.002] Defect ordering in aliovalently-doped cubic zirconia from first principles
Alexander Bogicevic, Christopher Wolverton, Gary M. Crosbie, Ellen B. Stechel (Ford Research Laboratory)
Defect ordering in aliovalently-doped cubic zirconia is
studied using gradient corrected density-functional
calculations. Intra- and inter-sublattice ordering
interactions have been investigated for both the cation (Zr
and dopant ions) and anion (oxygen ions and vacancies)
species. For yttria-stabilized zirconia, the crystal
structure of the experimentally identified ordered compound
Zr_3Y_4O_12 (\delta) is established, and we
predict the existence of new low-temperature zirconia-rich
ordered phases. Anion vacancies repel each other at short
separations, but show an energetic tendency to align as
third nearest neighbors along \langle 111 \rangle
directions. Doping with divalent (Be, Mg, Ca, Sr, Ba, Pt)
and trivalent (Y, Sc, B, Al, Ga, In) oxides shows that anion
vacancies prefer to be close to undersized cations, and away
from oversized cations. With intermediately sized cations,
the vacancies show no particular preference, and are thus
less prone to be tied up by the dopants when traversing such
oxides. This offers some simple insight into the high
conductivity of Y and Sc doped zirconia, as well the recent
success using, e.g., lanthanide oxides. Our calculations
highlight In as a particularly promising dopant/co-dopant
for high ionic conductivity.
[E18.003] Structural Form of Cubic BC_2N
Hong Sun (Department of Physics, University of California at Berkeley, Berkeley, CA 94720 and Department of Physics, Shanghai Jiao Tong University, Shanghai 200030, China), Soung-Hoon Jhi, D. Roundy, Marvin L. Cohen, Steven G. Louie (Department of Physics, University of California at Berkeley and Lawrence Berkeley National Laboratory, Berkeley, CA 94720)
Super-hard cubic boron-carbonitides (c-BC_2N) are
studied with the use of the ab initio psuedopotential
density functional method. Total energy, lattice constant,
bulk and shear moduli are calculated for all possible
c-BC_2N structures starting from the eight-atom cubic
zinc-blende structured unit cell. The results obtained
provide a plausible explanation for recent experimental
observations as well as a possible path to the synthesis of
the materials.
[E18.004] First-principles phonon dispersion and structural energetics of NiTi
Xiangyang Huang, Karin Rabe (Rutgers University)
NiTi undergoes a martensitic transition from a cubic B2
structure at high temperature to a low temperature
monoclinic B19' structure. We have investigated the low
energy surface of NiTi using first-principles total-energy
and density-functional-perturbation theory calculations in
the ultrasoft pseudopotential implementation of Baroni et
al. For cubic NiTi, we have computed the phonon dispersion
relation throughout the first Brillouin zone. The most
important features are instability for a small range of
wavevectors along the \Gamma-M line, and a more pronounced
instability at M. In a four atom supercell, we have computed
the energy for freezing in the M phonons, with and without
strain relaxation. In addition, the symmetry-breaking strain
which produces the observed ground state structure lowers
the energy still further, as expected. Our results for the
structural energetics for distortions of the cubic structure
are summarized in a polynomial energy expression, which
forms the foundation for effective Hamiltonian studies of
the martensitic phase transition and properties of the
high-temperature cubic structure.
[E18.005] Materials by Design: Cybersteels
Hern-Jeng Jou (QuesTek Innovations LLC)
A systems approach to the computational design of materials as dynamic hierarchical structures has been pioneered by the multi-institutional Steel Research Group centered at Northwestern University. Integrating process/structure/property/performance relations, the approach combines materials science, applied mechanics and quantum physics. Materials and design services based on the methods, tools and databases developed are being commercialized by QuesTek, employing QuesTek's Computational Materials Dynamics (CMD) system. Predictive control of nanostructures provides efficient strengthening, optimal control of transformation plasticity achieves record toughness, and quantum mechanical predictions provide Quantum Steels with enhanced intergranular cohesion and hydrogen resistance. A new class of ultrahard case-hardening steels with a 50surface hardness is entering a range of applications demanding high fatigue and wear resistance for greatly improved component life.
[E18.006] Automating First-Principles Phase Diagram Calculations
Axel van de Walle (Northwestern University), Gerd Ceder (MIT), Mark Asta (Northwestern University)
Although the formalism that allows the calculation of solid
state phase diagrams from first principles is well
established, its practical implementation remains a tedious
process. In order to maximize the technological impact of
this powerful tool, we have automated the process of
performing such calculations. An overview of the parameters
the algorithm requires as an input as well as its output is
given. The accuracy of the method is assessed through the
calculation of a variety of phase diagrams while its
usefulness in technological applications is illustrated
through the following example. It is well known that
disordered metallic alloys exhibit hardening in the presence
of short-range order. However, quantifying the level of the
short-range order from experimental measurements is
notoriously challenging, making it difficult to estimate the
correlation between short-range order and hardness. In
contrast, first-principles phase diagram calculations
directly provide very accurate short-range order parameters
at a fraction of the time and cost of accurate experimental
determinations.
[E18.007] Incorporating first-principles energetics in computational thermodynamics approaches
Chris Wolverton, Ravi Vijayaraghavan (Ford Motor Company), Xin-Yan Yan (U. of Wisconsin), Vidvuds Ozolins (Sandia National Labs.)
Computational thermodynamic approaches (e.g., those
pioneered by the CALPHAD community) have become a valuable
tool in the calculation of complex, multicomponent phase
equilibria often found in industrial alloys. These methods
rely on databases of free energies, which are often obtained
from an optimization process involving experimental
thermodynamic data and phase diagrams. However, many phases
of practical interest, such a precipitate phases, are
metastable, and the metastable phase boundaries are often
not well characterized. Consequently, these important phases
are often absent from computational thermodynamics
databases. We demonstrate that first-principles, density
functional calculations provide a means to obtain
thermodynamic functions of phases absent from current
databases. We illustrate this approach with the famous
metastable Cu-containing precipitate phases (GP zones and
Al_2Cu-\theta') often found in age-hardened aluminum
alloys. We discuss issues of the accuracy of
first-principles energetics and the incorporation of
absolute vs. relative energies into thermodynamics
databases.
[E18.008] Chemical Order of HCP Ag_2Al: resolving the discrepancy between X-ray and TEM data.
Nikolai Zarkevich, D. D. Johnson, A. V. Smirnov (Depts. of Materials Science amp; Engineering and Physics, Univ. of Illinois at Urbana-Champaign, Urbana, IL 61801), R. Hyland (K-B Alloys, Reading, PA)
Structural formation energies from electronic-structure
calculations are used to construct an effective alloy
Hamiltonian and then Monte Carlo simulations are performed
to study chemical ordering in HCP bulk and nano-precipitated
Ag-Al alloy. We find that the ground-state structure of bulk
HCP Ag_2Al is that proposed by Neumann(J.P.
Neumann, Acta Metall. 14), 505 (1966) from x-ray data,
while the structure of small HCP Ag_2Al precipitates in
Al matrix looks like those observed by TEM(J.M.
Howe et al.), Phil. Mag. A 56, 31 (1997) (with
Ag wetting the interface between matrix and precipitate).
These results resolve a long-standing apparent discrepancy
between X-ray diffuse scattering and TEM data and their
interpretation.
[E18.009] First-principles investigation of the rhenium effect in Cr-based alloys.
N.I. Medvedeva (Institute of Solid State Chemistry, Yekaterinburg, Russia and Northwestern University), Y.N. Gornostyrev (Institute of Metal Physics, Yekaterinburg, Russia), A.J. Freeman (Northwestern University)
While chromium exhibits attractive features like high melting temperature, high modulus to density ratio, and good oxidation resistance, the principle problem which limits its application is room temperature brittleness. The observed improvment of its ductility by alloying with Re was investigated with full-potential LMTO calculations including GGA corrections. The electronic structure and the ground-state properties (lattice parameters, elastic constants, cohesive energies) were calculated for bcc Cr and Cr-Re alloys within a wide range of Re concentration. The crystal parameters of assumed A15 and L1_2-type Cr_3Re compounds were optimized and their elecronic structure was compared with that for Cr-Re 25 % -alloy. The appearance of a metastable A15-phase for Cr-Re system was predicted and the conditions for its stabilization were analyzed. The mechanism of a possible ductilizing Re-effect appears to be the formation of small close-packed clusters or precipitates with A15-type structure which scavenge carbon impurities and avoid the formation of brittle chromium carbides.
[E18.010] Multiscale Modeling of Precipitate Microstructure: A Combined First-Principles/Cluster-Expansion/Phase-Field Approach
Venu Vaithyanathan, Long-Qing Chen (Penn St. Univ.), Chris Wolverton (Ford Motor Company)
Although highly accurate for predicting alloy properties,
current computational resources limit density functional
calculations to relatively small systems with a few hundred
atoms. Therefore, the direct application of these techniques
to problems of alloy microstructure (involving billions of
atoms or more) is clearly impossible. On the other hand,
continuum phase-field models have been successful at
accurately modeling alloy microstructure evolution; however,
these methods are formulated in terms of empirical or
difficult-to-measure thermodynamic input: 1) bulk free
energies, 2) interfacial energies, and 3) elastic strain
energies. Using the industrially-important problem of
Al_2Cu-\theta' evolution in aluminum alloys, we
demonstrate how first-principles atomistic calculations may
be combined with a mixed-space cluster expansion approach to
yield all of the necessary thermodynamic input for a phase
field model. The incorporation of these energetic
properties, obtained from atomistics, into a continuum
microstructural model represents a real breakthrough in
modeling capabilities: the first-ever ``first-principles''
model of precipitate microstructure evolution.
[E18.011] Effects of anisotropy on microstructural evolution in polycrystalline systems
Andrei Kazaryan, Bruce R. Patton, Yunzhi Wang (Ohio State University)
The important role of anisotropy in determining the structure and properties of systems as diverse as IC interconnects, magnetic recording materials, and nanocrystalline structures has led to increased interest recently in understanding how interface energy and mobility anisotropy control the formation of structures at the micro and nano scales. Advances in computer simulation of microstructural evolution, which allow incorporation of grain boundary anisotropies and rigid body motion of grains are reviewed*. Recent extensions of these models show that these properties can dramatically change behavior of polycrystalline systems and can be successfully used for the design of materials with unique structural and physical properties.
*A. Kazaryan, Y. Wang and B.R. Patton, Scripta
Mater.,vol.41,p.487 (1999); A. Kazaryan, Y. Wang S.A. Dregia
and B.R. Patton, Phys. Rev. B, vol.61, p.14275
[E18.012] Simulations of Dry Sliding Friction in a Model Lennard-Jones Solid
T.C. Germann, J.E. Hammerberg, B.L. Holian (Los Alamos National Laboratory)
We report on two-dimensional large-scale Molecular Dynamics simulations for Lennard-Jones solids, with initially flat commensurate and incommensurate interfaces, as a function of sliding velocity for various compressions. The high velocity behavior is similar for both commensurations, exhibiting an inverse power law for the frictional force when the velocity exceeds approximately one tenth the sound speed. At lower velocities, the frictional behavior differs, with the incommensurate case exhibiting a frictional force linear in velocity at low velocities followed by a maximum related to a phase transformation with a phase front propagating normally to the original interface. The frictional force for the commensurate interface approaches a finite value at low velocities.
[E18.013] Grain Boundary Dislocations and Sliding
John Hamilton, Istvan Daruka, Jonathon Zimmermann, Douglas Medlin (Sandia National Laboratory, Livermore, CA)
We describe modeling of grain boundary sliding and partial
dislocations at an aluminum sigma 3112 boundary. A
striking result of these first principles calculations is
that sliding the boundary results in a net ongoing
translation of the boundary in a direction perpendicular to
the boundary. First principles calculations were used to
determine the two stable symmetry-equivalent configurations
of this boundary which exist far from the partial grain
boundary dislocations. First principles nudged elastic band
calculations were then used to determine the Peierls
barriers associated with displacing the grain boundary
between these two symmetry-equivalent configurations.
Finally, a generalized Peierls-Nabarro Model was used to
obtain the partial dislocation widths from the calculated
Peierl's barriers. The dislocation core widths are wide due
to the small Peierls barriers. The theory is in good
agreement with new TEM observations of dislocation widths
presented here.
[E18.014] First principles study of dislocations in bcc metals
Darren Segall (MIT), Sohrab Ismail-Beigi (UC Berkeley), T.A. Arias (Cornell University)
The microscopic origins of plasticity in bcc structural metals are far more complex and less well understood, as compared to that of fcc and hcp metals. BCC metals have many active slip planes and violate the Schmid law. Understanding the microscopic origins of plastic flow in these metals are therefore crucial. We present new results on the structure and dynamics of the <111> screw dislocation in tantalum, which controls the low temperature plastic behavior of the material. Results in larger cells confirm our previous ab initio result[1] which cast doubt on the prevailing broken symmetry model for the dislocation core used to explain the non-Schmidt plastic behavior of bcc metals. We shall present these results, and the results of a new study of the critical yield stress for plastic flow in bcc tantalum.
[1] ``Ab initio study of screw dislocations in Mo and Ta: a new picture of plasticity in bcc transition metals,'' by Sohrab Ismail-Beigi and T.A.~Arias, Physical Review Letters 84:7, 1499--1502 (14 February 2000). Preprint: http://xxx.lanl.gov/abs/cond-mat/9908110~. http://xxx.lanl.gov/abs/cond-mat/9908110