R.K. Lawrence

Hole and electron trapping in ion implanted thermal oxides and SIMOX

Optically assisted methods of injecting either electrons or holes into SiO/sub 2/ layers were used to determine the effect of ion implantation on charge trapping in oxides. Dry-grown thermal oxides and the buried oxides of material grown by the SIMOX (separation by implantation of oxygen) process were studied. Al, Si, and P ions were implanted into the oxides at doses of 1/spl times/10/sup 13/ to 1/spl times/10/sup 16/ and the oxides were annealed at 700, 900, or 1050/spl deg/C after implantation. High dose implantations were found to create electron traps having high capture cross sections, the density of which depends on the implant species, suggesting that electron trapping is related to chemical aspects of the implanted ion. This was supported by measurements on an oxide implanted with a large dose of Ar, which showed no increase in electron trapping. It was found that the shift in the flatband voltage resulting from hole trapping could be reduced by high dose implantations, and that this effect is only weakly dependent on implant species. The hole trapping results are explained in terms of the effect of implantation on the oxide structure.

Annealing of total dose damage: redistribution of interface state density on [100], [110] and [111] orientation silicon

The annealing of interface states after an X-ray dose of 10 Mrad (SiO/sub 2/) under 1-MV/cm bias is studied on [100], [110], and [111] silicon. During annealing the bias is 1 MV/cm. Annealing times range from under an hour to hundreds of hours, and the temperature ranges from 75 to 175 degrees C. Using charge pumping, the energy distribution of interface states within the bandgap is determined. After annealing, the shape of the interface-state density curve implies that one or more defects other than P/sub b0/ and P/sub b1/ are present. Comparison of the interface-state density curves before and after annealing shows that a redistribution of interface-state density occurs over a large portion of the time-temperature range studied. The density near 0.4 eV above the valence band decreases and the density near 0.7 eV increases although the average density does not significantly change. Based upon the time scale and activation energy of the redistribution, a model is proposed in which the rate-limiting step is water diffusion within the gate oxide to the interface. This model provides a framework for a transformation among interface defects that accounts for the observed redistribution. Further tests for this model are discussed. >

H/sup +/ motion in SiO/sub 2/: incompatible results from hydrogen-annealing and radiation models

Two models that incorporate the same mobile H/sup +/ entity are the radiation model by McLean [1980] and the mobile charge model of hydrogen-annealed oxide by Vanheusden et al. [1997]. Mobile charge in hydrogen-annealed silicon-on-insulator (SOI) buried oxides before and after irradiation was studied to investigate discrepancies between the two models. We examined Unibond, low-dose SIMOX, and single- and triple-implant standard-dose SIMOX as well as single- and triple-implant SIMOX with supplemental oxygen implantation. To measure H/sup +/ motion as fast as 0.01 sec we developed a gate pulse method and combined it with standard I-V techniques to measure the full range of H/sup +/ response times. All but single-implant SIMOX exhibit mobile H/sup +/ that can be cycled between the Si/SiO/sub 2/ interfaces without reacting or being trapped. Trapping near the top interface attenuates the cycling of H/sup +/ in single-implant SIMOX. The transit H/sup +/ times were strongly affected by defects in the oxides and varied by an order of magnitude in oxides with the same thickness. The transit time varied linearly with oxide thickness. The effects of irradiation on the mobile H/sup +/ was studied to see if the irradiation would introduce defects that modify the H/sup +/ behavior and that bring the two models into agreement. No convergence was observed, After irradiation, H/sup +/ could be cycled between the Si/SiO/sub 2/ interfaces without reacting and its transit time across the oxide was not altered.

The role of nanoclusters in reducing hole trapping in ion implanted oxides

At the 2000 IEEE Nuclear and Space Radiation Effects Conference, it was shown that the negative shift in flatband voltage that results from hole injection is reduced in oxides that have been implanted with large doses of Al, Si, or P ions. In the present paper, we study the basic mechanism responsible for this reduced shift in the flatband voltage in more detail by comparing electron and hole trapping in Si and Ar implanted oxides. We find that in Si implanted oxides, the reduction in the shift of the flatband voltage is accompanied by the formation of entities in the oxide that have a large electron capture cross section, and that can become positively charged by photoemitting electrons. Photoluminescence studies indicate that these entities are Si nanoclusters. Oxides implanted with large doses of Ar do not form clusters, and these oxides show neither a reduction in the shift of the flatband voltage nor the formation of large capture cross-section electron traps. We show evidence that the nanoclusters reduce the shift of the flatband voltage by trapping protons formed during hole injection.

Reduction of radiation induced back channel threshold voltage shifts in partially depleted SIMOX CMOS devices by using ADVANTOX/sup TM/ substrates

Excessive total dose radiation induced back channel threshold voltage shifts often observed in fully depleted and partially depleted NMOS transistors fabricated in full dose SIMOX wafers can be greatly reduced by use of new low dose ADVANTOX/sup TM/ substrates.

Radiation induced charge in SIMOX buried oxides: lack of thickness dependence at low applied fields

Commercially prepared Separation-by-IMplantation-of-OXygen (SIMOX) wafers having buried-oxide (BOX) thicknesses ranging from 80 to 400 nm, and thermal oxides of similar thicknesses, were exposed to various doses of 10 keV X-rays. The net positive charge trapped in the buried-oxides during the radiation was determined by dual capacitance-voltage (C-V) and point-contact current voltage (I-V) measurements. For the thermal oxides, using metal-oxide-semiconductor (MOS) CV techniques, the amount of trapped charge was found to have the expected linear dependence on oxide thickness. For the SIMOX buried oxides, however, the amount of net trapped charge was found to be independent of BOX thickness when the oxides were biased at 0.05 MV/cm (a typical operating field). The SIMOX results are explained in terms of bulk-oxide hole and electron trapping.

The use of spectroscopic ellipsometry to predict the radiation response of SIMOX

We have studied SIMOX (Separation by Implantation of Oxygen) material using spectroscopic ellipsometry to determine the structure of the buried oxide and C-V measurements to determine the radiation response of the buried oxide. Our ellipsometric measurements indicate that the buried oxide is best described as a layer of stoichiometric SiO/sub 2/ which is more dense than bulk vitreous (v-) SiO/sub 2/. We find that the radiation response of the buried oxide is determined primarily by its density. We also find that small variations in the conditions used to prepare the SIMOX wafer can significantly affect the oxide density and its radiation response. The density of the buried oxide is also found to affect how it etches. >

A study of the radiation sensitivity of non-crystalline SiO/sub 2/ films using spectroscopic ellipsometry

Dry thermal oxides were grown on Si substrates at temperatures from 900/spl deg/C to 1200/spl deg/C. Portions of these oxides were then annealed for 10 min at temperatures from 900/spl deg/C to 1125/spl deg/C. Spectroscopic ellipsometry was used to determine the density of these oxides, and to look for the existence of an interfacial layer. It was found that the oxide density depends on both the growth temperature and the anneal temperature. No evidence was found of an interfacial layer. The oxides were then irradiated with X-rays, and the radiation-induced shift of the flatband voltage was determined. It was found that the radiation response was related to the oxide density determined by spectroscopic ellipsometry. These results suggest that the macroscopic structure of the oxide influences the extent of charge trapping.

Dependence of radiation induced buried oxide charge on silicon-on-insulator fabrication technology

X-rays have been used to irradiate material fabricated using various Silicon-on-Insulator (SOI) technologies, including the Bond and Etchback SOI (BESOI) process and several different Separation-by-Implantation-of-Oxygen (SIMOX) processes. The thickness of the buried oxide (BOX) in these samples ranged from 80 nm to 400 nm. The irradiations were done under high field (1 MV/cm) bias conditions. For each of these materials the net number of occupied traps in the buried oxide and the location of the charge centroid was determined. It was found that the location of the charge centroid depends on the density of the BOX, and on the radiation dose. It was also found that for buried oxides which have densities similar to that of fused silica, the net number of occupied traps in the BOX saturates at approximately 1.1/spl times/10/sup 13/ cm/sup -2/, and does not depend on the BOX thickness or fabrication technique. Both of these findings are important in device design, especially for substrates with thin buried oxides.

Charge trapping versus buried oxide thickness for SIMOX structures

Radiation induced charge trapping versus Buried-Oxide (BOX) thickness on various Separation-by-IMplantation-of-OXygen (SIMOX) buried oxides has been determined. An inflection point has been observed in the voltage shift versus buried oxide thickness relationship. As such, the radiation-induced voltage shifts for thin buried-oxides are greater than what could be expected from a simple square-law relationship. However, for irradiation applied fields higher than that of typical buried-oxide fringing fields the thickness relationship obeys a square-law. These results can be explained by the location and magnitude of the radiation induced oxide charge centroid and its relationship to the BOX thickness. The location of the centroid for trapped positive charge is dependent on the radiation-induced hole mobility, which is related to SIMOX processing, as well as on geometry and charge saturation.