Recent reports in the literature have suggested that Si2N2O forms in t
he oxidation of Si3N4 as a buffer suboxide below the silica crust, and
that equilibrium between SiO2 and Si3N4 requires the presence of this
buffer. Here we report the examination of SiO2/Si3N4 boundaries of di
fferent genesis, by a variety of techniques, all of which failed to de
tect Si2N2O. What was found in each case is a graded suboxide whose co
mposition merges seamlessly with the higher oxide above and the Si3N4
below. Part I presents the results of compositional depth profiling ac
ross the suboxide. In Part II a model is proposed to explain how O-2 d
iffusion in the graded suboxide limits Si3N4 oxidation kinetics.