Nature 436, 829–832 (2005)

This Letter described several notable physical properties for low-doped ruthenium copper oxides, from which conclusions concerning superconductivity in copper oxides were drawn. A key result was the observation of an unusual, negative, expansion of the lattice parameters and volume at temperatures below the Ru spin-ordering transition, TRu, as shown in Figure 3. Unfortunately, we have discovered that a discrepancy in the algorithm for fitting the diffraction data led to erroneous shifts in the cell parameters and resulted in the apparent negative expansion. We retract the claim of negative lattice expansion and the consequent inferences concerning the existence of two states in the pseudogap regime of the copper oxide electronic phase diagram. However, an anomalous expansion of the separation between copper oxide planes below TRu, as shown in Figure 3b, is reproducible and has subsequently been observed in other ruthenium copper oxide samples. Our results concerning large magnetoresistances and successive Ru and Cu spin-ordering transitions are not affected, so Figures 1 and 2 and the Supplementary Figures remain valid.

 We thank S. Kimber and A. Williams for undertaking further calculations to determine the nature of the fitting problem.