Passivation of III–V surfaces with crystalline oxidation

Abstract

Control of interfacial physicochemical properties associated with device materials to minimize the impact of point defects on device performance has been a dominant theme in the semiconductor industry. Control of the density of such defects for silicon has been well established for metal oxide-semiconductor field-effect device applications through deliberate reactions with chemically congruent species, such as hydrogen. In contrast, control of interfacial defects for technologically important III–V device materials is still an active area of research. Performance criteria for III–V devices are demanding in terms of energy efficiency, material consumption, sensitivity, and speed. The surface reactions of III–V crystals, including oxidation, are typically known to result in performance limitation for devices, causing significant degradation due to high defect-level densities at the surfaces/interfaces, in contrast to high quality bulk crystal regions. Here, we discuss the approach of utilizing atomically thin, ordered oxide interfacial layers of III–V compound semiconductors since they provide a unique opportunity for metal-oxide semiconductor applications, compared to the more common approach to avoid surface oxidation. Long-range ordered oxide interfaces have been obtained by oxidizing cleaned III–V surfaces intentionally in ultrahigh vacuum conditions. This can be combined with different passivation methods to decrease interfacial defect density in III–V devices. We present the current understanding of the physical and chemical properties of crystalline oxidized III–V materials, based on both experimental and computational models. The results are compared to those obtained by current state-of-the-art passivation methods.