The previous blogs of this series describe the failures that can occur on ICs for aerospace applications:
- SEL (single event latchup), described in Failures in Aerospace Applications, Part 2
- SEU (single event upset), described in Failures in Aerospace Applications, Part 3
- SET (single event transient), described in Failures in Aerospace Applications, Part 4
- SEB (single event burnout), described in Failures in Aerospace Applications, Part 4
- SEFI (single event functional interrupt), described in Failures in Aerospace Applications, Part 4
- SEGR (single event gate rupture), described in Failures in Aerospace Applications, Part 5
All the listed failures are due to the creation of e– /p+ because of the presence of radiation sources (proton/heavy ion) that impact the IC surface and cross the material. To determine the product's sensitivity to the heavy ions or to the protons, and to compare the products among themselves, main parameters have been defined:
- LET (MeV/mg/cm2 ): Linear energy transfer is the amount of energy transferred to the material by the impacting heavy ion/proton.
- LET threshold : This is the minimum starting LET from which SEE may occur on a component. The higher the threshold is, the better the component is.
- Fluence (#/cm2 ): It is the number of ions per unit area that hit the component.
- Cross-section (cm2 ): For a fixed LET , it is the ratio between the number of events detected on the device and the fluence .
- Saturated cross-section (cm2 ): From a certain LET , the number of events occurring on the components will reach a saturation level. The saturated cross-section is the maximum number of events divided by the fluence .
All these parameters are summarized on the Weibull curve (see Figure 1).
cross-section, and saturated cross-section.
The listed parameters (LET, LET threshold, cross-section, fluence, and saturated cross-section) are very useful to help the engineering team evaluate the robustness of the silicon technology which constitutes the IC to the displacement damages (DD).
There are two main sources of DD: protons and neutrons. These particles are massive compared to the atoms contained in the ICs, so they can create modifications in the structure of the IC as they pass through. The neutrons are neutral particles that don't generate any e– /p+ pairs, but the neutrons generate only DD. The protons are charged particles, which can thus generate both SEE and DD. All electronic components are composed of N or P doped silicon, which has internally free electric charges that can easily circulate inside these structured crystal lattices. If a neutron or a proton passes through the crystal structures, it disorganizes the crystal, and the disorganized electric charges don’t circulate anymore (see Figure 2).
through an N doped silicon material.
An IC may be more or less sensitive to the DD, depending on the internal structure of the IC. Let's consider, as an example, a N-channel MOSFET transistor. When the N-MOS is properly biased, the electrons flow from the source to the drain, inside a non-doped area, that is insensitive to the displacement damages, due to the absence of free charges (see Figure 3).
In another widely utilized integrated power switch, the IGBT, the situation is different. The conduction of the electrical charges happens in doped areas that may be affected by DD.
Did you ever evaluate the radiation performance of an IC for an aerospace application by means of its Weibull curve? What do you think of the effectiveness of this curve in comparing some products for aerospace applications? Have you ever experienced a DD on an IC?