Computational modeling the electrocaloric effect for solid-state refrigeration
- PDF / 202,359 Bytes
- 4 Pages / 432 x 648 pts Page_size
- 111 Downloads / 213 Views
Computational modeling the electrocaloric effect for solid-state refrigeration J.A. Barr1, T. Nishimatsu2, and S.P. Beckman1 1 Department of Materials Science and Engineering, Iowa State University, Ames, IA 50010, U.S.A. 2 Institute for Materials Research (IMR), Tohoku University, Sendai, 980-8577, Japan ABSTRACT The electrocaloric effect holds promise for possible application in refrigeration technologies. There is much interest in this subject and experimental studies have shown the possibility for creating materials with a modest sized electrocaloric response. However, theoretical studies lag behind the experimental effort due to the lack of computational methods to accurately study the finite temperature response. Here the freely distributed feram, an effective Hamiltonian molecular dynamics method, is demonstrated for predicting the electrocaloric response of BaTiO3. INTRODUCTION A pyroelectric crystal experiences a spontaneous change in polarization as its temperature changes.1 The electrocaloric effect (ECE) is the converse of this, i.e., the crystal experiences a spontaneous change in temperature when it's polarization changes. If the crystal is ferroelectric then the greatest pyroelectric and electrocaloric response occurs at the Curie temperature, where the crystal transforms from ferroelectric to paraelectric. It is possible to cycle the temperature and applied electric field to drive the conversion between thermal and electrical energies. One possible application is ECE based solid-state technologies for refrigeration. Although there is great interest in seeing the development of ECE materials, the development of theoretical methods lags behind the experimental effort. First-principles molecular dynamics methods are not yet feasible, and thermodynamic modeling, such as the Ginzburg-LandauDevonshire approach, requires substantial data for fitting. Here an effective Hamiltonian method will be demonstrated using both a direct and indirect molecular dynamics approach for the archetypical perovskite BaTiO3 (BTO). METHODS Here we use a first-principles based effective Hamiltonian model given by
M
H
eff
+V
short
+V
coup,hom o
=
* dipole
2
({u}) +V
.
¦uα (R) + 2
R,α
elas,hom o
* M acoustic ¦ 2 (R) +V self ({u}) + V dpl ({u}) 2 R,α w α
(η1,..., η6 ) + V elas,inho ({w})
({u}, η1,..., η6 ) +V coup,inho ({u},{w}) − Z * ¦ε •u(R).(1) R
39
This model follows that developed in Refs. 2,3. The collective atomic motion is coarse-grained by local soft-mode vectors u(R) of each unit cell at R in a simulation supercell. For further efficiency, local acoustic displacement vectors w(R) are not treated in the molecular dynamics, but are optimized according to u(R). The details of this model and the methods are given in Refs. 4,5. The molecular dynamics method is encoded in the feram software package, which is distributed free of charge under the GNU public license at Ref. 6. For the indirect approach the pyroelectric coefficient is predicted as a function of applied electric field and temperature. Then u
Data Loading...
