Electrical and physical characterizations of the effects of oxynitridation and wet oxidation at the interface of SiO₂/4H-SiC(0001) and

Abstract

The effects of oxynitridation and wet oxidation at the interface of SiO2/4H-SiC(0001) and were investigated using both electrical and physical characterization methods. Hall measurements and split capacitance–voltage (C–V) measurements revealed that the difference in field-effect mobility between wet oxide and dry oxynitride interfaces was mainly attributed to the ratio of the mobile electron density to the total induced electron density. The surface states close to the conduction band edge causing a significant trapping of inversion carriers were also evaluated. High-resolution Rutherford backscattering spectroscopy (HR-RBS) analysis and high-resolution elastic recoil detection analysis (HR-ERDA) were employed to show the nanometer-scale compositional profile of the SiC-MOS interfaces for the first time. These analyses, together with cathode luminescence (CL) spectroscopy and transmission electron microscopy (TEM), suggested that the deviations of stoichiometry and roughness at the interface defined the effects of oxynitridation and wet oxidation at the interface of SiO2/4H-SiC(0001) and .

Impact of the NO Anneal on the Microscopic Structure and Chemical Composition of the Si‐Face 4H‐SiC/SiO2 Interface

“…The O and C profiles match very well and show a gradual change from one bulk material to the other. There is no significant C excess in any of the measurements, which is in agreement with many recent studies . The value of w tr for each measurement was determined from both the O and the C profiles and is indicated by the hatched area.…”

“…A certain transition region of a few nanometers where the stoichiometry appears to gradually change from bulk SiC to SiO 2 has commonly been observed but the reported values of the width of the transition region ( w tr ) range from 1.5 nm up to 25 nm . Reported correlations between w tr and the channel mobility have not been confirmed in other studies . Those examples demonstrate the level of disagreement in the literature and the need to revisit this topic to establish more clarity which structural changes are relevant for the mobility improvement.…”

“…Furthermore, it has been reported that the N concentration at the interface is up to the equivalent of one atomic layer (≈1.2 × 10 15 cm −2 ) and confined within a layer of ≈1.5 nm . Even though such a concentration should be detectable by spatially resolved EELS, there is only a limited number of EELS studies that reported on a small N signal at the interface . It is emphasized that both of these studies only showed clearly resolved N peaks on C‐face interfaces, while a reliable and reproducible detection at the technologically more relevant Si‐face has not been attained with EELS yet …”

“…However, the main cause of the mobility improvement and the passivation mechanism of the NO anneals are still under debate. Numerous studies attempted to reveal the structure and chemical composition of the SiC/SiO 2 interface using scanning transmission electron microscopy (STEM) and electron energy‐loss spectroscopy (EELS) . However, there is no consensus on the influence of the NO anneal and which of the observable structural changes play a significant role in the mobility improvement.…”

“…The initially low mobility and the high D it in unpassivated devices has often been explained by C residuals from the oxidation or nonstoichiometric transition layers in the interface region . While in earlier EELS studies such transition layers have been reported, in most recent studies no such interlayers with an increased C content have been found . A certain transition region of a few nanometers where the stoichiometry appears to gradually change from bulk SiC to SiO 2 has commonly been observed but the reported values of the width of the transition region ( w tr ) range from 1.5 nm up to 25 nm .…”