Tetragonal tungsten bronze compounds: relaxor vs mixed ferroelectric - dipole glass behavior

Abstract

We demonstrate that recent experimental data (E. Castel et al J.Phys. Cond. Mat. {\bf 21} (2009), 452201) on tungsten bronze compound (TBC) Ba$_2$Pr$_x$Nd$_{1-x}$FeNb$_4$O$_{15}$ can be well explained in our model predicting a crossover from ferroelectric ($x=0$) to orientational (dipole) glass ($x=1$), rather then relaxor, behavior. We show, that since a "classical" perovskite relaxor like Pb(Mn$_{1/3}$ Nb$_{2/3}$)O$_3$ is never a ferroelectric, the presence of ferroelectric hysteresis loops in TBC shows that this substance actually transits from ferroelectric to orientational glass phase with $x$ growth. To describe the above crossover theoretically, we use the simple replica-symmetric solution for disordered Ising model.