Posted by IkingJr: Wed Jun 16, 2004 8:27 am    Post subject: Eutectic Temperature for various die back metallizations

Hello EE;

Our team is confused, we recently did a eutectic die attach (die back - 100 wt %) on one of our package using ESEC 2005HR @ 420ºC bond temp. We were able to get very good and stable die shear readings (3.5 CpK) with almost 100% silicon remain on all samples. We were advised by demiGODS of the company that the standard "eutectic" temp is 440ºC, where did they get this?.
To what I know, eutectic temperature for Au-Si systems (eutectic composition of 97.1 Wt % Au : 2.85 Wt % Si.) is at 363ºC only - is this correct? How about for gold back only? Can someone clarify this issue and explain how to get the right "eutectic temperature" based on different die back metallizations.

Regards,

Posted by EE: Fri Jun 18, 2004 3:48 pm    Post subject:

Hi Iking Jr.,

There's nothing wrong with using a D/A temp of 420 deg C if you're getting superior results from this temp. I've seen some companies doing eutectic Au-Si D/A at a temperature even lower than this, e.g., 410 deg C.

This is not to say, however, that you should refuse the bosses' request to do it at 440 deg C. After all, this is still data. Make a well-planned evaluation to compare the outputs of the process at these two temps, and come up with a good report stating your final recommendation.

EE

Posted by IkingJR: Mon Jun 21, 2004 11:30 am    Post subject:

Dear EE;

Thanks. Actually our group is very confident that we can operate below 440ºC with superior result and we have demonstrated that in our report. To further justify our stand, can you recommend a known or standard reference that we can use showing eutectic temp of different back metallization types. Is there such a standard?

Best regards,
Eric

Posted by Newman: Sat Sep 18, 2004 1:23 am    Post subject:

Hi ALL,

Just to help AU-SI die attach temp can vary also depending on the total time needed to bond all dies and also on the inert gaz flow (if argon 90/10) and of course if you perform or not the scrubbing when bonding ).

From 98 we were optimising the eutectic temp per type of dice using using the non destructif testing SONOSCAN Equipment and not the die-shear testing .

For sure the GOLD back side metallisation is impacting the die-bonding integrity and quality.

We have modified more than 20 SPECS ""EUTECTIC TEMP"" depending on the % of voids and the electrical test results .

RF power assembly engineer Morocco.
nbadidi@---

Posted by Guest: Sun Oct 23, 2005 10:55 am    Post subject: Re: Eutectic Temperature for various die back metallizations

 
The temperature mentioned is given by customer. Its the same on most of their packages around the subcon. Their reason is in the variation of gold thickness at the wafer back and this temperature as tested by them could produce good result on this variation.

Posted by Sebastian: Wed Nov 09, 2005 1:38 pm    Post subject: Re: Eutectic Temperature for various die back metallizations

 
Some inputs too...

Are you doing eutectic DA on a hermetic package? PGA? Side brazed? If not, nevermind this reply.

If so, you need to check not only the adhesion of the die to the pad via die shear test, or what temperature setting you have in your machine but also for loose particles.

Hermetic packages have cavities (not molded) that these loose materials (usually remnants of the Au/Si preform during scrubbing) do not adhere much on the die pad. During centrifuge test, these get detached and become loose particles inside the cavity. You need to perform PIND (particle induced noise detection) test to check for these particles.

And these particles usally get "loose" because of higher temperature setting and prolonged scrubbing time during eutectice DA.