Oxidation precipitates in silicon

In a recent paper [J. Appl. Phys. 50, 7996 (1979)] Leroy discussed the kinetics of growth of oxidation stacking faults. It is here suggested that it is unlikely that a self-interstitial stacking fault on {111} in a diamond structure crystal would have the form proposed by Leroy, since it does not preserve even nearest-neighbor bonding. An alternative structure, which preserves both nearest- and next-nearest-neighbor relationships, is described. It is further suggested that, since self-interstitials in silicon and germanium are known to aggregate on the low-energy {113}, any model for oxidation precipitates on {111} which assumes a simple self-interstital growth mechanism is unlikely to be correct.