It is demonstrated that cuprous oxide (Cu2O) can be electrodeposited epitaxially onto silicon (Si) and indium phosphide (InP) (001) single‐crystalline substrates from aqueous solution. Epitaxial electrodeposition under these conditions is remarkable considering the strong driving force for the formation of amorphous native oxide layers on Si and InP substrates. To elucidate the growth mechanisms, the microstructure of the interfaces between the Cu2O layer and the two different substrates was investigated by TEM (transmission electron microscopy) in conjunction with XEDS (X‐ray energy‐dispersive spectroscopy) and EELS (electron energy‐loss spectroscopy). In both heteroepitaxial systems, the Cu2O layers have a unique but non‐trivial crystallographic orientation relationship (OR) with the substrate, which can be described as a 45° rotation around the common [001] axis representing the substrate normal. We show that this relationship minimizes the overall misfit between corresponding interatomic spacings of the two adjacent crystals. In apparent contradiction to the unique OR, TEM revealed that in both hetero‐systems the Cu2O layer is separated from the substrate by an amorphous interlayer. The thickness of the interlayer typically is a few nanometers. The presence of an amorphous interlayer contrasts with our experimental results on electrodeposited Cu2O on Au (001) single‐crystal substrates, also included in this article, where TEM shows the Cu2O epilayer in direct contact with the substrate. XEDS and EELS analysis of the chemical composition and bonding at Cu2O/Si and Cu2O/InP interfaces in the as‐grown state as well as after tempering revealed that the interlayer is mainly composed of SiO2 and InPO4, respectively. Most likely, the observed epitaxial layers on top of an amorphous interlayer evolve by nucleation of epitaxial Cu2O directly on the substrate. While simultaneous oxidation of the substrate leads to the formation of an amorphous layer, the epitaxial nuclei can laterally overgrow the oxide. Consequently, the local composition of the amorphous layer varies with the nature of the substrate.

Immersion in aqueous hydrofluoric acid ͑HF͒, a standard processing step in fabricating microelectromechanical systems devices, causes rougher and weaker polycrystalline silicon ͑polysilicon͒ films in the presence of Au metallization and leads to relatively thick surface oxides. The origin of these effects is the galvanic corrosion that takes place due to the difference in electrochemical potential between the noble Au metallization and the polysilicon in HF solution, resulting in the accelerated formation of a porous surface layer due to corrosion of the silicon. The electrochemical behavior and the evolving microstructural morphology are correlated. In HF solution, Si is dissolved and porous silicon ͑PS͒ is formed. The PS first develops along the polysilicon grain boundaries and then extends into the grain interiors. After rinsing in water and exposure to air, the pore walls form the usual oxide, leading to a Si/SiO 2 composite with a gradient in composition-from a high fraction of SiO 2 near the surface to 100% Si deeper into the polysilicon. Inasmuch as the galvanic corrosion requires positively charged holes, these effects can be minimized by using n-type dopants in the polysilicon and performing the HF immersion in the absence of illumination.

No abstract

No abstract

No abstract