"Wigner crystal" and "stripe" models for the magnetic and crystallographic superstructures of La0.333Ca0 .667MnO3

TL;DR

This study investigates the charge and magnetic superstructures of La0.333Ca0.667MnO3 using high-resolution diffraction, supporting a Wigner crystal model over stripe models, and refines the magnetic and charge orderings at low temperatures.

Contribution

The paper provides experimental evidence favoring the Wigner crystal model for charge ordering in La0.333Ca0.667MnO3, challenging the stripe model based on bulk diffraction data.

Findings

Charge ordering involves Mn3+ orbital ordering and superlattice peaks.

Wigner crystal model explains superlattice peak intensities.

Stripe model is inconsistent with bulk diffraction data.

Abstract

The crystallographic (charge-ordered) and magnetic superstructures of La0.333Ca0.667MnO3 were studied by high-resolution synchrotron x-ray and neutron powder diffraction. In the antiferromagnetic structure, which was refined using a non-collinear model, the a lattice parameter is tripled and the c lattice parameter is doubled with respect to the average crystallographic unit cell (Pnma setting). The crystallographic structure below the charge-ordering temperature (TCO ~ 260 K) is characterized by ordering of the dz2 orbitals of the Jahn-Teller-distorted Mn3+O6 octahedra in the orthorhombic ac plane, and the appearance of superlattice peaks in the x-ray patterns corresponding to a tripling of the a axis lattice parameter. The intensities of the superlattice peaks can be accounted for satisfactorily in terms of ordering of the Mn3+ cations in sites as far apart as possible in the ac plane ("Wigner crystal" model) and transverse displacements of the Mn4+O6 octahedra in the c direction. These results are not consistent with a recently proposed model based on transmission electron microscopy (TEM) data in which the Mn3+O6 octahedra are arranged in a stripe pattern ("stripe" model). In particular, the large longitudinal stripe modulation revealed by TEM is not observed, suggesting that the TEM data may not be representative of the bulk sample. Within the framework of the "Wigner crystal" model, the magnetic structure at 1.5 K and the charge-ordered superstructure at 160 K and 1.5 K were refined from the neutron data.