Fluctuating Defects in the Incipient Relaxor K_1-xLi_xTaO_3 (X=0.02)

C. Stock,P. M. Gehring,G. Xu,D. Lamago,D. Reznik,M. Russina,J. Wen,L. A. Boatner

DOI: https://doi.org/10.1103/physrevb.90.224302

2015-01-01

Abstract:We report neutron scattering measurements of the structural correlations associated with the apparent relaxor transition in K1-xLixTaO3 for x = 0.02 [KLT(0.02)]. This compound displays a broad and frequency-dependent peak in the dielectric permittivity, which is the accepted hallmark of all relaxors. However, no evidence of elastic diffuse scattering or any soft-mode anomaly is observed in KLT(0.02) [J. Wen et al., Phys. Rev. B 78, 144202 (2008)], a situation that diverges from that in other relaxors such as PbMg1/3Nb2/3O3. We resolve this dichotomy by showing that the structural correlations associated with the transition in KLT(0.02) are purely dynamic at all temperatures, having a time scale on the order of similar to THz. These fluctuations are overdamped, nonpropagating, and spatially uncorrelated. Identical measurements made on pure KTaO3 show that they are absent (within experimental error) in the undoped parent material. They exhibit a temperature dependence that correlates well with the dielectric response, which suggests that they are associated with local ferroelectric regions induced by the Li+ doping. The ferroelectric transition that is induced by the introduction of Li+ cations is therefore characterized by quasistatic fluctuations, which represents a stark contrast to the soft-harmonic-mode-driven transition observed in conventional perovskite ferroelectrics such as PbTiO3. The dynamic, glasslike structural correlations in KLT(0.02) are much faster than those measured in random-field-based lead-based relaxors, which exhibit a frequency scale of order similar to GHz and are comparatively better correlated spatially. Our results support the view that static random fields give rise to the relaxor phenomena, and that the glasslike dynamics observed here characterize a nascent response.