On the surface analysis of copper oxides: the difficulty in detecting Cu3O2

Abstract

The existence of Cu3O2, a gross defect structure of Cu2O, has been documented experimentally since the early 1960s. However, discussions of the oxidation of copper often neglect the importance of this phase; in fact, it is often omitted entirely from such discussions. This results from the difficulty in determining the chemical state during sputter depth profiling and relying on techniques that have difficulty providing chemical state information. The occurrence of sputter reduction during the depth profiling of copper oxide layers has been demonstrated with XPS depth profiles on a series of copper samples oxidized, for varying lengths of time, in air at a temperature of either 423 or 523K. Under these conditions, a thin layer of CuO/Cu(OH)2 terminates the oxide layers. Beneath this layer, the presence of Cu3O2 is expected on the samples prepared at 423K. However, immediately upon the beginning of sputtering, only Cu1+ is detected in the oxide layers. A zone of constant Cu:O ratio of (approximately ∼1.5) is found throughout most of the oxide layer even though Cu2+ is not detected. On the samples prepared at 523K, the presence of CuO is anticipated. However, Cu2+ is not detected after sputtering is initiated and a region of constant Cu:O ratio of ca. 1.5 is detected. The inherent difficulties involved in investigating oxide layer growth and vertical oxide layer structure using sputter depth profiling are discussed in light of this experimental data.