Comment on ``Comparison of the $Cc$ and $R3c$ space groups for the superlattice phase of Pb(Zr$_{0.52}$Ti$_{0.48}$)O$_3$''

Abstract

The two-phase (space groups $Cc$ and $Cm$) structural model recently proposed for the modeling of the neutron powder diffaction pattern collected on Pb(Zr$_x$Ti$_{1-x}$)O$_3$ (PZT) sample with $x\approx 0.52$ [R. Ranjan, A. K. Singh, Ragini, and D. Pandey. \textit{Phys. Rev. B} \textbf{71}, 092101 (2005)] is shown to neglect the $hkl$ dependent (anisotropic) line broadening. The most serious problem with this model is that it assigns octahedral tilts to a wrong phase. Instead of correctly taking the anisotropic line broadening into account, this model used low symmetry phases to minimize the residuals during Rietveld refinement. The essential features of a model taking the $hkl$ dependent line broadening into account are summarized. It has already been reported that once the anisotropic line broadening, revealed by high resolution neutron powder diffraction instrument, is correctly taken into account the model with $R3c$ and $Cm$ space group symmetries describes all Bragg peaks and their intensities well [J. Frantti, S. Eriksson, S. Hull, V. Lantto, H. Rundlöf, and M. Kakihana. \textit{J. Phys.: Condens. Matter} \textbf{15}, 6031 (2003).]. It was further shown that the $Cm+R3c$ model is consistent with the structural features observed at other compositions and temperatures, which is particularly important in the vicinity of the phase boundary. Also problems related to sample preparation and data collection are pointed out.