DEMOCRITOS - ACTIVITIES

Piezo

PHYSICS OF PIEZOELECTRIC AND FERROELECTRIC MATERIALS (Piezo)

Activity Coordinator:

Raffaele Resta

Participants:

Agnese Callegari

Collaborators:

R. Dovesi

X. Gonze

F.Mauri

N. Marzari

Main research lines:

Dielectric polarization: macroscopics and microscopics (Raffaele Resta)
Localization and local quantities in electronic structure (Raffaele Resta)
Solid-Solution materials (Stefano Baroni)

Piezoelectric and ferroelectric materials play an overwhelming role in the technology of sensors and other devices: they are therefore of widespread use in everyday's life. Despite this fact, the theoretical modelling and understanding of these materials has lagged behind, and until recently was based on phenomenological/empirical approaches only. Many fundamental issues remained unexplained.

In the last decade there have been sweeping advances in the theory of these materials, owing to the development of new ideas and algorithms, based on first-principle electronic structure. These advances would not have been possible without a seedling effort by the U.S. Office of Naval Research (ONR), which steadily supported the leading theorists in the field, including the present Coordinator. In 2001 a grant by ONR (about one million dollar at start up) was used to establish the Center for Piezoelectrics by Design (CPD), http://www.cpd.wm.edu/, a dedicated facility whose sole non U.S. member is the present Coordinator. In Europe this same area of research remains marginal and underdeveloped, despite the importance of the materials and of their physical properties.

1 DIELECTRIC POLARIZATION: MACROSCOPICS AND MICROSCOPICS

The physical property which makes the present materials useful, namely macroscopic polarization, has evaded for many years even a precise microscopic definition, and has challenged first-principle calculations. Most textbooks contain incorrect definitions. Only ten years ago was the phenomenon analyzed in the correct way, and quantum-mechanical calculations of macroscopic polarization in real materials became possible. Owing also to the development of novel, physics-based, algorithms [1], several first-principle calculations appeared, providing fundamental insight into the phenomenon of dielectric polarization itself.

Macroscopic polarization is an integrated quantity: simply a vector in any homogeneous sample. What is microscopic polarization, and how is this qualitatively different in different materials? We plan to study the local current, i.e. the transient microscopic current which flows through the insulating material while polarization is switched on. This current is clearly at the origin of macroscopic polarization, but about nothing is known about it in any material. For instance, a material enjoying a high dielectric or piezoelectric constant can be thought of as one where the transient polarization current flows "easily". What does that mean, and how it happens, are central issues of this research. In particular the relationship of this microscopic current to atomic environment, coordination, and kind of bonding, will be investigated on some test-case materials.

2 LOCALIZATION AND LOCAL QUANTITIES IN ELECTRONIC STRUCTURE

First-principle studies of materials have been based, almost invariably, on wavefunctions: however, the wavefunction has a delocalized nature, while the physics is essentially local, particularly so in insulators [2]. For instance, often a given chemical bond has an environment-independent energy: nonetheless the computation of a bond energy requires the complete wavefunction of a possibly very large system. Methods for overcoming this drawback are starting to emerge.

Besides the local current discussed in line #1 of the present activity, we will pursue the study of other local quantities in insulators (dielectric, piezoelectric, and ferroelectric materials). In this class of local properties are densities (energy density, stress density), density matrices, localized orbitals (Wannier-like) [3], dynamical charges [4] and other "shielding factors", essential to understand the physics of dielectrics and piezoelectrics.

Studies of local quantities are computer experiments providing an unbiased insight, which goes far beyond the one accessible to related laboratory experiments. In an initial stage we will essentially focus on the a-posteriori analysis of wavefunctions produced in the standard way: later on, the knowledge acquired is likely to be used in a constructive way.

Localization and polarization characterize insulators and are, at a very fundamental level, two aspects of the same phenomenon [2]. Polarization and localization properties are a challenging benchmark for energy functionals within DFT [5]: as for the SurfInt activity, we expect significant developments in this area of general interest.

3 SOLID-SOLUTION MATERIALS

The most important class of ferroelectric and piezoelectric materials are the perovskite oxides ABO_3, where A is a mono- or di-valent cation, B is a penta- or tetra-valent cation (transition metal), and O is oxigen. The materials of interest, however, invariably come in the form of alloys: for instance, the most common piezoelectric in everyday's life applications is PZT (solid solution of PbZrO_3 and PbTiO_3). This material, as well as most technologically useful ones, is ceramic (i.e composed of randomly oriented crystallites) and cannot be synthetized in single-crystal form. The reasons for this fact are unknown and the present activity could shed some light on the phenomenon.

The field was revolutionized in 1997 when a new single-cystal solid-solution piezoelectric was discovered (acronym PMN-PT), displaying a giant electromechanical coupling [6]. Since then, a large theoretical effort has been devoted to understand the phenomenon with some success [7], but the key role of single-crystal morphology, as well as the one of compositional order, are still unknown. We believe that brute-force electronic structure calculations on large supercells are not the good means to attack alloy properties. Instead, based on the outstanding previous achievements of some CRS members in the theory of semiconductor solid solutions [8], we are confident that novel viable and elegant methods will be found and that new physics can be discovered in the present materials too.

INTERACTION WITH OTHER RESEARCH GROUPS

This activity will use codes and methodology partly common to other groups in the CRS (4 and 6, particularly in the use of DFT, PDFT, and TDHF), and will both lead to and benefit from the developments of novel energy functionals. The workers in this activity will have a strong interaction with some physicist in the U.S.: among them the CPD members, R. Car (Princeton), and N. Marzari (MIT). In Europe, CRS plans to become the reference center for promoting the physics of ferroelectricity and piezoelectricity, fostering the interactions among the (presently few) active workers in the field.

BIBLIOGRAPHY

[1] R. Resta, Rev. Mod. Phys. 66, 899 (1994); Europhysics News 28, 18 (1997); Phys. Rev. Lett. 80, 1800 (1998).
[2] R. Resta and S. Sorella, Phys. Rev. Lett. 82, 370 (1999).
[3] N. Marzari and D. Vanderbilt, Phys. Rev. B 56, 12847 (1997).
[4] R. Resta, J. Phys. Chem. Solids 61, 153 (1999).
[5] S.J.A. van Gisbergen et al., Phys. Rev. Lett. 83, 694 (1999).
[6] S.E. Park and T.R. Shrout, J. Appl. Phys. 82, 1804 (1997).
[7] H. Fu and R.E. Cohen, Nature 403, 281 (2000).
[8] S. de Gironcoli, P. Giannozzi, and S. Baroni, Phys. Rev. Lett. 66, 2116 (1991); N. Marzari, S. de Gironcoli, and S. Baroni, Phys. Rev. Lett. 72, 4001 (1994).

last modified: 03.06.2004


INFM crs DEMOCRITOS SISSA - Via Beirut 2-4 34014 Grignano, Trieste, ITALY Tel. +39 0403787443 Fax. +39 0403787528

webmaster: webadmin@democritos.it