Direct Observation of Charge Order and an Orbital Glass State in MultiferroicLuFe2O4

Abstract

Geometrical frustration of the Fe ions in LuFe2O4 leads to intricate charge and magnetic order and a strong magnetoelectric coupling. Using resonant x-ray diffraction at the Fe K edge, the anomalous scattering factors of both Fe sites are deduced from the (h/3 k/3 l/2) reflections. The chemical shift between the two types of Fe ions equals 4.0(1) eV corresponding to full charge separation into Fe2+ and Fe3+. The polarization and azimuthal angle dependence of the superlattice reflections demonstrate the absence of differences in anisotropic scattering revealing random orientations of the Fe2+ orbitals characteristic of an orbital glass state.

Strong local lattice instability in hexagonal ferritesRFe2O4(R= Lu,Y,Yb) revealed by x-ray absorption spe

“…the distribution should be Fe 2+δ with 0.25 < δ < 0.75 and not necessarily bimodal). This experimental result seems to contradict previous RXS studies [7,40] in LuFe 2 O 4 that were considered to support the existence of Fe 2+ /Fe 3+ CO. Let us comment shortly on the limitations of the reported RXS analysis. Satellite reflections (h/3, k/3, l/2) that appear in the low-temperature CO phase come from a strong nonresonant Thomson scattering plus the energy-dependent anomalous term.…”

“…Mulders et al [40] do not determine either the Thomson scattering or the α parameter that is given by the CO model and the crystallographic structure. Otherwise, they fit these two parameters as a unique parameter in their model and fix the chemical shift (charge disproportion).…”

“…On one hand, the claimed colossal dielectric constants are spurious effects originating from the electrical contacts [37][38][39], and on the other hand, LuFe 2 O 4 does not show spontaneous electrical polarization, as has been proved by electrical polarization measurements [39]. Although the occurrence of ferroelectricity can be discarded in LuFe 2 O 4 , the structural phase transition at T CO 320 K is still considered to originate from Fe 3+ /Fe 2+ CO. Resonant x-ray scattering (RXS) experiments at the Fe K-edge of the principal observed superstructure (1/3, 1/3, half-integer) reflections seem to support this assumption, but the analysis performed by either Ikeda et al [7] or Mulders et al [40] are merely qualitative, using a generic structure factor without considering any charge order sequence. Other authors have shown that the CO involves charged, rather than polar, bilayers with four different iron valences, but with values that are still grouped in a bimodal distribution of two different iron atoms, referred as Fe 3+ -and Fe 2+ -like [27].…”

“…Polycrystalline powders of RFe 2 O 4 (R = Lu,Y,Yb) were obtained by solid state chemistry reaction from stoichiometric amounts of R 2 O 3 and Fe 2 O 3 and sintered at 1200°C (R = Lu,Yb) and 1240°C (R = Y) in a CO 2 /CO (60: 40) atmosphere. Similarly, LuFeCoO 4 powder samples were prepared in air at 1350°C using Lu 2 O 3 , Fe 2 O 3 , and CoO as precursors.…”

Determination of the sequence and magnitude of charge order inLuFe2O4by resonant x-ray scattering

“…RXS experiments at the Fe K-edge were reported for LuFe 2 O 4 [5,35,36] and the energy and temperature dependence of the superlattice (1/3,1/3,l/2) reflections was used to support the occurrence of full charge segregation into Fe 2+ and Fe 3+ . Nevertheless, the analysis performed by either Ikeda et al [5] or Mulders et al [36] only used a generic structure factor with both the Thomson scattering and the resonant terms as free parameters in the fits without considering the low-temperature crystallographic structure and the CO sequence.…”