Electric field dependent thermal conductivity of relaxor ferroelectric PMN-33PT through changes in the phonon spectrum

In ferroelectric materials, an applied electric field has been shown to change the phonon dispersion relation sufficiently to alter the lattice thermal conductivity, opening the theoretical possibility that a heat gradient could drive a polarization flux, and technologically, also opening a new avenue toward all solid-state heat switching. In this report, we confirm experimentally the validity of the theory originally developed for Pb(Zr,Ti)O_3 (PZT) on the ferroelectric relaxor 0.66Pb[Mg_(1/3)Nb_(2/3)]O_3-0.33PbTiO_3 (PMN-33PT). In the theory, the relative change in sound velocity and thermal conductivity with applied electric field relates to the piezoelectric coefficients (d_33 and d_31) and the Gr\"uneisen parameter ({\gamma}). The theory predicts that in PMN-33PT the effect should be an order of magnitude larger, and of opposite sign as in PZT; this is confirmed here. The effect is measured on samples that undergo multiple field sweep cycles and pass through 2 phase transitions. The thermal conductivity changes are closely linked to variations in the piezoelectric coefficients. A null experiment on paraelectric SrTiO_3 confirms the link between ferroelectricity and phonon spectrum changes. Changes in heat conduction as large as 10% can be achieved over large temperature ranges, opening a possibility for practical heat switches based on phonon spectrum changes.
Comment: 19 pages, 2 tables, 8 figures