We have analyzed red and blue luminescence from porous silicon as a function of oxidation parameters and feature dimension determined with an atomic force microscope. We have found correlation between blue luminescence intensity and the increase in feature size caused by oxidation. We have further shown that blue luminescence, is identical, with respect to spectrum and fast decay, to that of high microelectronic quality SiO, grown on crystalline silicon using dry oxygen plus an organic chlorine compound. Thus, we conclude that blue luminescence originates from SiOZ film rather than from the silicon nanocrystals in the porous material. Intensity enhancement, as compared to SiO, on crystalline wafers, comes from the gigantic surface area of porous silicon.

The photodissociation of iron-boron pairs in p-type silicon produces lifetime killing interstitial iron and may be combined with noncontact surface photovoltage (SPV) measurement of the minority carrier diffusion length to achieve fast detection of iron. We found that, for iron concentrations ranging from 8×108 to 1×1013 atoms/cm3, the pair dissociation using a white light (10 W/cm2) was completed within 15 s. Surface recombination was a major rate limiting factor. Passivation of the surface enhanced the rate by as much as a factor of 20. The photodissociation rate increased with increasing temperature, however, the increase was smaller than that of the thermal dissociation rate. These characteristics are consistent with a previously proposed recombination enhanced dissociation mechanism. For practical iron detection, it is important that the detection limit of the approach is close to one part per quadrillion.

A procedure is presented for determining long minority carrier diffusion lengths, L, from the measurement of the surface photovoltage (SPV) as a function of the light penetration depth. The procedure uses explicit SPV formulas adopted for diffusion lengths longer than the light penetration depths. Results obtained on high-purity silicon demonstrate new capability for noncontact wafer-scale measurement of L values in a mm range, exceeding the wafer thickness by as much as a factor of 2.5. This factor can be increased by increasing the accuracy of SPV signal measurement. The procedure does not have the fundamental limitations of previous SPV methods in which the diffusion lengths were limited to about 70% of the wafer thickness.

No abstract

No abstract