Solid-state thermal-interface bondline: Silicon bonded directly to aluminum

[TPI_bondline]

What design and cost advantages could be gained, in what applications, with the bonding of silicon wafer -- the heat source in power devices -- directly to aluminum -- the heat sink..?

Despite the massive CTE-mismatch, we've shown that this Si-on-Al laminaion is possible, with the bondline being able to survive thermal shock and aging ... but haven't been able to quantify the advantages and/or identify targeted applications. It seems that electronic-packaging folks conservatively stick with the conventions of layered laminates to step down the CTE, from aluminum's 23 ppm/oC to silicon's 2.6 ppm/oC.

Thanks!

 

[3DDave]

Hard to believe no one is kicking your door down for this. What's the effect on electrical performance from very high and variable strain on the silicon?

 

[IRstuff]

bondline being able to survive thermal shock and aging

Lattice strain changes the bandgap, which drastically changes the performance of silicon as a semiconductor. The bondline was the least of the problems.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[3DDave]

It also occurs to me this will tend to put the rather brittle silicon into tension and the top surface is covered in stress raising notches.