http://www.electronicdesign.com/boards/rohs-implementation-challenges
see the cross section of Ceramic cap - metallization got Ni layer...  
not seen on your image.  hopefully, it is not MFG missing a process  
(or cheap stuff for low temp/ Ag adhesive potting type).
On Sep 6, 2017, at 3:01 PM, Yuan-chia Joyce Koo wrote:

> did it missing a Ni plating?  http://www.johansondielectrics.com/ 
> understanding-ceramic-capacitor-terminations
> best of luck.
> jk
> On Sep 6, 2017, at 12:25 PM, Giamis, Andy wrote:
>
>> Hi TechNetters,
>>
>> Here’s something I have not seen before.
>> I have seven different ‘theories’ for what’s happening here.
>> Three of which do not involve mythical creatures, illegal  
>> substances or the alignment of celestial bodies.
>>
>> This capacitor is on a path that sees a lot of RF power.
>> The unit went through 20 cycles of -70C to +115C while at nominal  
>> power.  Temperature sensing suggests this device could be hitting  
>> 170C.
>> No, we normally don’t test to these extremes.
>> The PCB has ENiG finish and the solder is SAC.   This is normal  
>> SMT, no rework.
>>
>> I think I know in general what happened, but I am surprised.
>> I don’t have an explanation for the formation of the big pocket on  
>> the right.
>> Has anyone ever come across any solder joints like this?
>>
>> Thanks Steve for uploading the pictures (links below).
>>
>> http://stevezeva.homestead.com/pockets_25.jpg
>>
>> http://stevezeva.homestead.com/Pockets_CS.jpg
>>
>> http://stevezeva.homestead.com/Pocket-left.jpg
>>
>> http://stevezeva.homestead.com/Pockets-right.jpg
>>
>>
>> Best Regards,
>> Andy
>> Andrew C. Giamis
>> Senior Failure Analysis Engineer
>> CommScope
>> 2601 Telecom Pkwy
>> Richardson, TX, 75082, USA
>>
>>