An analysis is performed of the dependence of lifetime on excess carrier density when traps are present. Using only realistic assumptions, a relatively simple solution is obtained. Different types of behavior can be obtained, depending on recombination parameters and the concentration and energy-level position of the traps. An especially interesting situation exists in which lifetime versus excess density curve shows no indication of trapping, yet incorrect recombination parameters are obtained if trapping is ignored. Experimental confirmation of the analytical results is presented.

An understanding of the charge-transport process in SiO2 is vital to an understanding of radiation effects in MOS devices. To account for hole transport in SiO2 below room temperature, a stochastic hopping transport model has been employed by others with considerable success. An alternate model is presented here which explains many experimental data in an even more satisfactory fashion. This model is based on multiple trapping at a continuum of trapping levels, the concentrations of which decrease exponentially with energy from the valence band edge. By adjusting the parameters of the model to fit the data at a single temperature, behavior at other temperatures for the same specimen is accurately predicted, including the variation of apparent activation energy with the amount of transport that has occurred. This feature and others which are discussed are not shared by the hopping model previously employed.

Carrier transport in SiO2 has been studied by exciting hole-electron pairs in an oxide film by a pulsed electron beam. The energy of the beam (4–8 kV) was chosen to minimize excitation in the Si substrate upon which the SiO2 was grown. Measurements of oxide current vs applied voltage were made with beam intensity and energy as parameters for SiO2 layers of three thicknesses. It is demonstrated that normalized current-vs-field curves are independent of beam intensity, beam energy, and film thickness over the range studied. These results indicate that the analysis used by various workers to determine the mobility-lifetime (μτ) product for SiO2 is invalid. Observed dependences of current on applied field can best be explained by geminate and/or columnar recombination. The present findings indicate that both holes and electrons traverse most of the SiO2 without appreciable permanent trapping. The amount of positive charge trapped at or near the SiO2–Si interface is a significant fraction of the ``collected'' charge, indicating that some of the holes that move to the interface do not penetrate it.

No abstract

Charge transport studies have been performed on SiO films using an electron-beam injection technique· MOS capacitors incorporating oxides grown at 1000 and 1100°C were investigated, including units fabricated at Hughes Aircraft using radiation hardening p r o c e dures. A comparison of beam-induced current v s field characteristics is made for devices with differing pro cessing h i s t o r i e s . Additionally, experimental deter minations of trapped positive charge vs collected charge were performed. Present findings indicate that holes are mobile in SiO£, that the schubweg model is insufficient for describing charge transport in S1O2 films, and that the electron-hole pair creation energy for S i 0 2 i s ^ 19 eV. Current vs field data can be qualitatively explained in t e r m s of columnar and/or geminate recombination. Conclusions concerning the effects of processing on charge buildup are made and a qualitative model based on experimental findings is presented. Implications of this model for radiation hardening are discussed.