Abstract
Potassium-tantalate-niobate (KTN) is an archetypal ferroelectric alloy whose atomic structure and phase transition mechanism have been debated for decades. The key to resolving this debate lies in understanding the off-centering displacements of transition metal (TM) atoms (Nb, Ta), which is essential to distinguish two competing scenarios of the ferroelectric transition: displacive vs. order–disorder. Here, we combine neutron scattering techniques that cover large energy and momentum spaces, with machine learning augmented molecular dynamics, to fully resolve the collective nature and disorder in the positions of TMs in KTN. This work settles down the longstanding questions about correlated atomic disorder in an exemplary ferroelectric, from which the useful functionality emerges.
