Guy,
the std BGA  need to pass either ball pull or ball share test... it  
should not separated from UBM... unless it is prototype unit...  
didn't go through QA process... (after thermal shock or rel cycle  
test, it could separated from UBM if crack initiation site present at  
UBM/substrate solder mask area).  just a guess.
jk
On Dec 21, 2017, at 4:30 PM, Guy Ramsey wrote:

> Thanks for the thoughtful response. One of the other MEs here is  
> thinking
> that this is a normal appearance, that the SAC305 solder uniformly
> separated from the intermetallic when we reflowed and pulled the BGA.
>
> I wish this was the case but it seems unlikely. Isn't the affinity  
> of the
> bulk solder to the intermetallic higher than its surface tension?
>
> On Thu, Dec 21, 2017 at 4:26 PM, Yuan-chia Joyce Koo  
> <[log in to unmask]>
> wrote:
>
>> Steve, your friend might need cross section and see if UBM is done
>> correctly, besides, the solder mask looks extra thick for a device...
>> didn't look like std BT substrate... cross section of good vs  
>> after reflow
>> one might give you some clue what is wrong... the color of the   
>> pad look
>> like poor adhesion of the nickel layer... it might be poorly done  
>> either
>> possiviation layer prior to balling (either ImAg, ImAu or ImTin), or
>> missing etchant step (or pH adjustment not done properly) prior to  
>> last
>> immersion plating step...IMHO.
>> On Dec 21, 2017, at 2:39 PM, Steve Gregory wrote:
>>
>> Hi all,
>>>
>>> I've been asked to post this to the Technet by a dear friend:
>>>
>>> *We installed this little processor. At test we identified opens  
>>> on the
>>> device. When we removed it we saw a number of pads that were  
>>> matte gray
>>> with no evidence of solder adhered to the pad. *
>>> *We attempted to bump the pads with solder, they did not wet  
>>> easily. In
>>> some cases we had to scrape through the matte finish to a shiny  
>>> metal.
>>> But, we did get all the pads to appear wetted.*
>>>
>>> *We cleaned the site well and then fluxed the bumps with a tacky  
>>> flux ROLO
>>> designed for POP, placed the BGA and reflowed it on a rework  
>>> station.*
>>>
>>> *The assembly failed test with similar result, opens under the  
>>> processor.*
>>>
>>> *The attached photo shows that the bump process only appeared to  
>>> wet the
>>> pads. The condition returned when we removed the BGA  second time.*
>>>
>>> * http://stevezeva.homestead.com/2017-12-21_11.21.17.jpg
>>> <http://stevezeva.homestead.com/2017-12-21_11.21.17.jpg>*
>>>
>>> *Have you seen this before? Is there any recovery? This assembly  
>>> is worth
>>> about $45K. *
>>>
>>> Thanks,
>>>
>>> Steve
>>>
>>> --
>>> Steve Gregory
>>> Kimco Design and Manufacturing
>>> Process Engineer
>>> (208) 322-0500 Ext. -3133
>>>
>>> --
>>>
>>>
>>> This email and any attachments are only for use by the intended
>>> recipient(s) and may contain legally privileged, confidential,  
>>> proprietary
>>> or otherwise private information. Any unauthorized use,  
>>> reproduction,
>>> dissemination, distribution or other disclosure of the contents  
>>> of this
>>> e-mail or its attachments is strictly prohibited. If you have  
>>> received
>>> this
>>> email in error, please notify the sender immediately and delete the
>>> original.
>>>
>>