Wide discrepancies exist among recent studies in the determination of the position of an oxygen adsorbate with respect to the Cu(100) substrate having the purported $c(2\mathrm{}\times 2)$ superstructure. Using surface treatments described in the recent literature and a sensitive low-energy electron diffraction detector, we find the oxygen superlattice is not a mixture of $c(2\mathrm{}\times 2)$ and $(\sqrt{2}\times 2\sqrt{2})R45$ superstructures but has only a single ordered overlayer, namely the $(\sqrt{2}\times 2\sqrt{2})R45$ structure. This result contradicts the conclusions of recent reports.

• Received 12 December 1985

DOI:https://doi.org/10.1103/PhysRevB.33.8899