Posted by Sadoc: Thu Dec 07, 2006 9:22 pm    Post subject: FA for open issue

hello all:

I have some question for the fail mode for IC open issure.
The fail mode as below:

1.under the x- ray , no abnormal be observed. Gold ball still on the pad,
and wire weep is ok. But in fact , the IC is open.

2.after decap for the ball shear test or the neck pull test, it always fail
pull value is nearly zero,same with shear.

Could anyone tell me why?

Posted by FARel Engr: Thu Dec 07, 2006 10:08 pm    Post subject:

Hi Sadoc,

Can you please describe the intermetallic coverage (IMC) on the affected bond pads? If the IMC's are insufficient, then you likely have a contam on the surface of your bond pads that prevents good intermetallic formation. If so, you need to have the bond pads analyzed with EDX to determine the nature of the contaminants.

Good Luck!

FARel Engr

Posted by Sadoc: Thu Dec 07, 2006 10:40 pm    Post subject:

Thanks for your reply.
OK, well i don't really sure the IMC percentage, but there are some au-al left on the pad, i guess may be for 70% or higher.

I can show you the EDX atomic confirm result
1.test on the left gold area ,C 29.51%,O:32.67%,AL:10.43%,AU:27.39%
2.test on the pad ,C:20.46%,O:4.71%,AL:67.56%,Si:2.83%,Ti:2.44%

SO, how about the O good or bad for the bondability?
and the how about the C, is it show up for the normal or abnormal?

Posted by FARel Engr: Thu Dec 14, 2006 10:01 pm    Post subject:

Hi Sadoc,

If 70% IMC is left on the bond pad, then that means you have very good intermetallic formation, so surface contam may not be your issue. You must then ask yourself why the bonds are lifting even if the IMC's are so good. Is there a possibility that you've overbonded them?

The high % of C and O are a bit abnormal. Also, if you are detecting the Ti metal under the bond pad, then you are doing the EDX analysis at a very high EHT. Please do the EDX analysis at a much lower EHT, say 5 keV so that what you'll get are data from the surface of the bond pad, and nowhere else. Do the EDX analysis at 10 keV as well for comparison.

BTW, what is your bond pad metal's composition? Is it doped with Si? Why are you seeing Si in the bond pad itself?

FARel Engr

Posted by Logan218142: Sat Dec 16, 2006 4:10 pm    Post subject:

Hi... do we know which point you detected 'open' fail, is this a time zero fail from SMT line or reliability or field failure?

On PFA side, is there any SAM analysis or X-section done? Any specific site/pads having 'open connections'

Do you observe ductile structure or micro-voids of IMC on 'open' bonding pads?