Running a chip at 100C for an hour should not affect the long term ESD protection.

I will admit that the ESD protection level is a function of operating temperature. However, most people will try to avoid touching equipment that is running at 100C.

If you are using epitaxial silicon to prevent SCRs, then the ESD level is set by the layout of the input / output buffer. There is a nasty phenomena called secondary breakdown that allows an output buffer to concentrate all of the ESD current in a very small region (<1u). This causes localized heating, which can exceed the eutectic temperature of aluminum and silicon. And that is a very bad thing.

You limit this by increasing the drain resistance, thereby swapping one bad thing for another.

--------

I'm not sure about corrosive gas. But I would imagine that the effects would look similar to metal migration failures, which are a second major factor in chip lifetime. Electrons tend to knock aluminum atoms out of the way, and over time the metal opens up. This is strongly affected by temperature, and so extended temperature operation will definitely reduce the life of your product.

For example, we're investigating making some 150C-200C processors for seismic work, and we would like to achieve a 1000 hour (6 week) lifetime. We may settle for less. And we will probably double the width of all metal lines to increase the chip lifetime.