First-principles Studies of the Local Structure and Relaxor Behavior of Pb(Mg1/3Nb2/3)

Hengxin Tan,Hiroyuki Takenaka,Changsong Xu,Wenhui Duan,Ilya Grinberg,Andrew M. Rappe
DOI: https://doi.org/10.1103/physrevb.97.174101
2018-01-01
Abstract:We have investigated the effect of transition-metal dopants on the local structure of the prototypical $0.75\mathrm{Pb}({\mathrm{Mg}}_{1/3}{\mathrm{Nb}}_{2/3}){\mathrm{O}}_{3}\text{\ensuremath{-}}0.25{\mathrm{PbTiO}}_{3}$ relaxor ferroelectric. We find that these dopants give rise to very different local structure and other physical properties. For example, when Mg is partially substituted by Cu or Zn, the displacement of Cu or Zn is much larger than that of Mg and is even comparable to that of Nb. The polarization of these systems is also increased, especially for the Cu-doped solution, due to the large polarizability of Cu and Zn. As a result, the predicted maximum dielectric constant temperatures ${T}_{\mathrm{m}}$ are increased. On the other hand, the replacement of a Ti atom with a Mo or Tc atom dramatically decreases the displacements of the cations and the polarization, and thus, the ${T}_{\mathrm{m}}$ values are also substantially decreased. The higher ${T}_{\mathrm{m}}$ cannot be explained by the conventional argument based on the ionic radii of the cations. Furthermore, we find that Cu, Mo, or Tc doping increases the cation displacement disorder. The effect of the dopants on the temperature dispersion $\mathrm{\ensuremath{\Delta}}{T}_{\mathrm{m}}$, which is the change in ${T}_{\mathrm{m}}$ for different frequencies, is also discussed. Our findings lay the foundation for further investigations of unexplored dopants.
What problem does this paper attempt to address?