Re-examination of the room temperature crystal structure of CuGeO3 by x-ray diffraction experiments: Observation of new superlattice reflections

A recent electron paramagnetic resonance (EPR) study of CuGeO3 revealed the existence of an antisymmetric exchange interaction, Sigma(i) d(ii+1) . (S-i x S-i+1), between Cu spins on the c-axis in this compound (Yamada I, Nishi M and Akimitsu J 1996 J. Phys.: Condens. Matter 8 2625). To allow this interaction, the crystal structure must have no inversion centre halfway between relevant Cu sites. However, the structure with the space group Pbmm (D-2h(5)), which was proposed for this compound by Vollenkle et al and has been accepted so far, does not allow this interaction because it has an inversion centre at the midpoint between nearest-neighbour Cu sires on the c-axis. This conflict was pointed out in the EPR study and was taken to indicate the inconsistency of the structure with Pbmm. To find the correct crystal structure of CuGeO3, x-ray diffraction experiments are performed at room temperature on samples which are markedly improved in their quality because they are produced by the annealing and slow cooling of single crystals grown by the Boating-zone method. As a result, new superlattice reflections which had not been reported so far are detected. The space group and the unit cell are determined to be P2(1)2(1)2 (D-2(3)) and 2a(p), x b(p) x 4c(p), respectively, where a(p) x b(p) x c(p) is the pseudo-unit cell proposed by Vollenkle er al. The newly found structure allows the antisymmetric interaction mentioned above. Possible connections between the new structure and the question of why the structural transition. which brings about the dimerization of Cu spins, occurs in CuCeO3 are discussed.