IC product lifetime as function of junction temperature

If an IC is rated for an absolute maximum junction temperature of, say, 170°C, obviously it is not recommended to operate there - but how drastically is product lifetime impacted if we are close, say operating at a junction temperature of 160°C? How severely does an IC's lifetime get shortened as we get closer to the maximum junction temperature?

There are two rules of thumb when it comes to premature aging of electronics and temperature:

1. Every 10°C above 25°C halves its life

2. Every 15°C above 25°C halves its life.

The 10°C is derived from a certain application of Arrhenius' equation

$$AF = e^{ \frac{E_a}{k}}\left(\frac{1}{T_\mathrm{use}}- \frac{1}{T_\mathrm{test}}\right)$$

The issue with this is the 10°C result was a very broad interpretation of the empirical results (no consideration was given to other failure modes).

MIL-HDBK-217 took into account field data and concluded that 15°C is a figure more applicable to practical usage

https://www.electronics-cooling.com/2017/08/10c-increase-temperature-really-reduce-life-electronics-half/

• I think that Every 10°C above 25°C halves its life is related to chemistry. Commented Nov 24, 2018 at 11:27
• I agree, this was then broadly applied. I have been looking into this recently wrt uprating. The problem is the mail is out of date
– user16222
Commented Nov 24, 2018 at 13:16
• And to just put a number to it. $2^{\frac{160-25}{15}}=512$. That means, if the device has a lifetime of 512 years at 25° C, then it will have a lifetime of 1 year at 160° C. Commented Nov 25, 2018 at 2:49

Should be half its life with every 10°C above 125°C; with Tj = 165°C, life time would be about 1 or 1.5 yr; also, plastic package may be damaged quickly.

• This is a pretty old answer, but I would love to see a link to literature for this. Commented Apr 27, 2023 at 21:12