Estimating and Reducing Errors in MOS Dosimeters Caused by Exposure to Different Radiations

This paper is concerned with the errors that can result because of charge recombination differences in the SiO2 layer of a PMOS transistor that is to be employed as an integrating dosimeter aboard an earth-orbiting satellite. The space radiation environment consists mostly of electrons and protons of various energies. These electrons and protons will produce tracks with different ionization densities in the SiO2. Consequently there will be differences in the amount of charge recombination, and these differences will be reflected as errors in the dose measurements. Data have been obtained on the devices for different radiations. Saturation curves relating the shift in threshold voltage, ¿VT, per krad (SiO2) as a function of gate bias during irradiation, VG, are obtained for devices exposed to Co-60 gamma rays, electrons with a most probable energy of 7 MeV, 37-MeV protons, and Cu X rays. A renormalization procedure is applied to the saturation curves in order to remove from consideration the confusing effects caused by systematic errors in the dosimetry. The resulting set of curves then permit the determination of the errors in dose measurement associated with only charge recombination. From these data the allowable range of gate bias is determined that will result in reasonably accurate measurements of dose in the space radiation environment. For a given choice of gate bias, the set of renormalized curves also yields an objective estimate of the possible error associated with differences in charge recombination.