S.M. Prokes, O.J. Glembocki, V.M. Bermudez, R. Kaplan, L.E. Friedersdorf, P.C. Searson, "SiHx excitation: An alternate mechanism for porous Si photoluminescence," Physical Review B, 45:13788-91, 1992.
Sharka M. Prokes (Naval Research Laboratory, Washington, D.C.): "Silicon has many attractive properties, such as abundance, a stable passivating oxide, and good thermal and mechanical properties, which make it one of the most important materials in the electronics industry. Unfortunately, bulk silicon is a very poor optoelectronic material because its electronic structure is such that only inefficient light emission in the infrared can be produced. This occurs because the optical process is second-order in nature, involving both photons and phonons.
"Porous silicon is produced by electrochemical etching, and differs from bulk silicon in that it consists of interconnected silicon structures, some on the order of several nanometers. The discovery of strong, visible luminescence (approximately 1.7 eV) from porous silicon generated significant interest, and it was believed that the recombination of electrons and holes within these nano-crystallites was the source of this light emission (L.T. Canham, Applied Physics Letters, 57:1046, 1990). If correct, this would make it possible to tune the color of light emission by choosing the appropriate silicon particle sizes.
"The study by our group was the first to show that the light-emission energy was dependent on the presence of surface species, such as hydrides or oxyhydrides, and not on the particle size. A competing model for the light-emission process was suggested, based on the surface properties of this material, and independent of particle size. Interestingly, more recent work (Y. Kanemitsu, et al., Phys. Rev. B, 48:2827, 1993) showed that the 1.7 eV light emission does not correlate with the silicon particle sizes. Furthermore, high-temperature laser heating experiments (S.M. Prokes, et al., Phys. Rev. B, 49:2238, 1994) have shown the 1.7 eV light emission to be temperature-independent, indicating that the emission process cannot be a property of the silicon crystallites, but may be a property of oxide-related color centers present at the surface of the silicon nanostruc-tures."