Research on Prediction Method for Heavy Ion SEE Cross-section Based on Proton SEE Cross-section

Abstract

With the development of space technology and semiconductor technology, more and more microelectronic devices are used in spacecraft. Heavy ions and protons are two main particles that cause single event effect (SEE) in devices. Heavy ions cause SEE mainly through direct ionization. In most cases, protons have too low linear energy transfer (LET) to cause SEE directly, and they induce SEE through the ionization of the secondary ion products created by nuclear reactions between protons and device materials. The relationship between proton and heavy ion SEE physical mechanism gives us a chance to calculate proton SEE cross-sections from heavy ion experiment data and vice versa. Most of past researches have focused on predicting proton SEE cross-sections from heavy ion SEE cross-sections. However, there are few studies on predicting heavy ion SEE cross-sections based on proton ones. At present, there are the following problems in heavy ion SEE experiments. Firstly, the device needs to be decaped or background before irradiation, which costs time and money, especially for new flip-chip devices, it is difficult to remove the package because of their complex structure. Secondly, due to the short range of heavy ions in the air, the use of low energy heavy ions in experiments requires vacuum environment, but vacuum pumping and breaking consume much time. The proton SEE experiments don’t have these problems, so it is necessary to carry out the research on the method of predicting heavy ion SEE cross-sections from proton ones. Here, a method was proposed to predict heavy ion SEE cross-sections based on proton ones. The theoretical relationship between heavy ion and proton SEE cross-sections was obtained, proton SEE cross-sections can be obtained by integrating heavy ion ones with equivalent LET spectra, and the calculation method for predicting the heavy ion SEE cross-sections was established based on their theoretical relationship. The key of the calculation method is the equivalent LET spectrum of deposited energy of proton incident device. The problem of calculating heavy ion SEE cross-sections from proton SEE cross-sections is transformed into the problem of solving the equation, whose unknows are the four parameters of Weibull function which describe the change of heavy ion SEE cross-sections with LET, and uses particle swarm optimization (PSO) algorithm to solve the problem. Three static random-access memories (SRAMs) with different feature sizes was selected to verify this method, and the results are in good agreement with their heavy ion SEE experimental data. This method provides a new way to obtain heavy ion SEE cross-sections of microelectronic devices, especially for flipchip devices.

Keywords

• single event effect
• heavy ion
• proton
• linear energy transfer
• particle swarm optimization algorithm