Weak charge-lattice coupling requires reinterpretation of stripes of charge order in (La,Ca)MnO3

Authors:
  • JLJ. C. Loudon
    J. C. Loudon
  • SCS. Cox
    S. Cox
  • AWA. J. Williams
    A. J. Williams
  • JAJ. P. Attfield
    J. P. Attfield
  • PLP. B. Littlewood
    P. B. Littlewood
  • PMP. A. Midgley
    P. A. Midgley
  • NMN. D. Mathur
    N. D. Mathur

Abstract
Manganites of appropriate composition display modulations attributed to stripes of charge, orbital and spin order. It is believed that the charge order arises when Mn valence electrons are localised by a strong electron-lattice coupling that locks together the superlattice and undistorted parent lattice periodicities. For example, the doubled periodicity observed in La0.5Ca0.5MnO3 is ascribed to orbitally ordered alternating planes (stripes) of the idealised cations Mn3+ and Mn4+. The expected lock-in of periodicities implies that (chemically tuned) arbitrary levels of charge doping should generate stacking faults in the charge order. Therefore one expects La0.48Ca0.52MnO3 to possess alternating Mn3+ and Mn4+ planes, with an extra sheet of Mn4+ ions every 7.8 nm. Surprisingly, electron diffraction patterns taken at 90 K with a convergent beam of 3.6 nm diameter never display the doubled periodicity expected between stacking faults. Instead, we record a uniform non-integer periodicity with no stacking faults. This discovery is supported by our finding that large area electron diffraction peaks are sharp compared with simulations that assume such stacking faults. Given that electron-lattice coupling is unable to lock together the two periodicities, one is forced to reconsider whether it is large enough to localise electronic charge strongly in modulated manganites of the form La1-xCaxMnO3.