Several issues have been raised concerning the validity of an interfacial nonbridging-oxygen-hole-center (NBOHC) mechanism suggested for the red photoluminescence (PL) observed in porous silicon. These issues include the PL energy range, the effects of total oxygen content on the PL intensity, and the expected absorption process. It is shown that the expected PL energy range for the NBOHC-type emission is in the 1.4–2.0-eV range, due to changes in local chemistry and strain; the red PL intensity is not expected to scale in any simple way with total oxygen content, but it does scale remarkably well with the presence of oxygen-related shallow donors, as measured by electron-spin resonance; and the absorption process (obtained by excitation spectroscopy) in porous silicon shows a remarkable similarity to the absorption process obtained for the NBOHC in silica glasses. That is, both result in an absorption in the 2- and 4.8-eV ranges, which is quite different from that of silicon. Thus no inconsistencies exist between the NBOHC-type model and results obtained in as-made or oxidized porous silicon. Furthermore, in view of the significant similarities between the optical behavior of porous silicon and that of oxide-related centers, the NBOHC-type mechanism cannot be discounted as the source of the red PL in porous silicon, for both as-made and oxidized samples.

• Received 11 October 1994

DOI:https://doi.org/10.1103/PhysRevB.51.11183