Photoluminescence mechanism model for oxidized porous silicon and nanoscale-silicon-particle-embedded silicon oxide

There is much debate about the photoluminescence (PL) mechanisms of the nanoscale Si/Si oxide systems containing oxidized porous silicon and a nanoscale-Si-particle (NSP)-embedded Si oxide deposited by chemical vapor deposition, sputtering, or Si-ion implanting into Si oxide. In this paper, we suggest that two competitive processes, namely, the quantum confinement (QC) process and the quantum confinement-luminescence center (QCLC) process, take place in the PL. The photoexcitation occurs in the NSPs for both of the processes, while the photoemission occurs either in the NSPs for the QC process or in the luminescence centers (LCs) in Si oxide adjacent to the NSPs for the QCLC process. The rates of the two processes are compared quantitatively. Which process plays the major role in PL is determined by the capture cross section, the luminescence efficiency, and the density of the LCs, and the sizes of the NSPs. For a nanoscale Si/Si oxide system with the LCs having certain capture cross-section and luminescence efficiency, the higher the LC density and the larger the sizes of NSPs, the more beneficial for the QCLC process to surpass the QC process, and vice versa. For certain LC parameters, there is a critical most probable size for the NSPs. When the most probable size of the NSPs is larger than the critical one, the QCLC process dominates the PL, and when the most probable size of the NSPs is smaller than the critical one, the QC process dominates the PL. When the most probable size of the NSPs is close to the critical one, both the QC and QCLC processes should be taken into account. We have used this model to discuss PL experimental results reported for some nanoscale Si/Si oxide systems.