Simulating the effects of the space ionizing radiation environment (total dose) on microcircuits

The near earth space ionizing radiation environment consists of continuous but time varying proton and electron belts and transient radiation from solar activity (mostly protons) and deep space (mostly gamma rays and protons). The energy and flux of these ionizing particles varies widely. Ionizing radiation may cause severe degradation of the parametric electrical performance of microelectronic devices and in some cases even functional failure. Degradation of bipolar transistors in space was first observed in the Telstar satellite in 1962 after a high altitude nuclear event increased the electron flux in the belts. Ionization induced degradation of Metal Oxide Semiconductor (MOS) field effect transistors was reported as early as 1964. Hence the effects of ionizing radiation on microelectronic components have been known for over forty years. Early studies showed that to simulate the "effects" of space ionizing radiation with ground testing it is not necessary to use the same types, energies and fluxes of the space ionizing particles. The reason is that the "effect" of the radiation is proportional to the absorbed ionizing energy or dose. The dose is measured in rad, where 1 rad equals 100 ergs/gm, or gray, where I gray equals 1 joule/kg. For many years it was accepted that the damage to microelectronic devices was semi-permanent and was only a function of the dose. Test standards were based on this concept. Starting in the 1970s it was recognized that reality was more complex. In MOS devices it was shown that the physical processes leading to the degradation were time dependent and occurred over a time scale of many decades both during and after the application of the radiation. Hence the rate at which the radiation was received (flux or dose rate) was important. An accelerated test technique was developed to simulate the effects of the low dose rate space environment with a higher dose rate laboratory environment by using a combination of irradiation and post irradiation annealing. The test standards were modified to include these more complex tests. While the modifications worked well for MOS devices it was shown in the early 1990s that things were even more complex for many bipolar linear circuits. For these circuits there is a "true" dose rate effect that can not be simulated with the test developed for MOS devices. It has been more difficult to develop accelerated tests to simulate the low dose rate space environment for bipolar linear circuits and work continues to optimize such a test. Hence the early concept of simulating the space ionizing radiation environment with a single parameter, dose, has proven to be far too simplistic and several modifications to the total dose test standard have been required.