Agree with George 100%.  if you use Pb/Sn  solder and Pd containing  
finishing, you risk for brittle IMC in plate format... not sure about  
Pb-free... the Pb/Sn reaction with Pd formation of IMC is very fast  
if my memory serve me right.
jk
On Jan 22, 2018, at 6:39 PM, George Wenger wrote:

> Hi Guy,
>
> I'm a really old school soldering person.  There is nothing better  
> than SnPb soldering to an electroplated copper surface.  Also I  
> still hate the "Black Pad" name.  Black Pad isn't a failure mode it  
> is a name given to what you see after a failure.  Black Pad  
> failures have been evaluated for years and I still don't think they  
> really have pinned down the cause for the failure mode.  In my 45  
> years in the telecommunications industry the one thing I learned is  
> that solder joints made to electroless platings are not as strong  
> as solder joints made to electrolypic platings.  The attached file  
> (sorry TN will strip it but you will see it) lists some of the  
> brittle interface failures I've experienced during my years doing  
> failure mode analysis.  The common factor is solder joints made to  
> electroless platings are prone to brittle interface failures.   
> Although I've listed the manufacturer in the attached list the  
> failures are not due to the manufacturer but to the type of surface  
> finish. Other than immsersion tin which and ENIG , which I really  
> dislike, EPENIG is my other least favorite surface finish.  It has  
> three negatives (Pd and two electroless platings EP and EN).  Pd as  
> a top coating on a PCB does what Pd in a cars catalictic converter  
> does (i.e., it attacks crap out of the environment it is in) and it  
> is easily contaminated and loses it's solderability.  If you don't  
> dissolve all of the EPd you solder to an electroless Pd surface and  
> if you do you solder to an Eni  surface and those solder joints are  
> not strong and subject to brittle interface failure.
>
> I've always been a fan of IAg.  Yes silver oxidizes and yes silver  
> tarnishes but most of  the fluxes used in soldering can reduce the  
> oxide and tarnish and if the tarnish is really bad you know it by  
> sight before you do the soldering.  And the silver is so thin it  
> dissolves into the solder and you solder to the electroplated  
> cooper below the immersion silver so you have as strong a solder  
> joint as possible.
>
>
> So I'm not surprised that when you are using components with ENIG  
> plating and PCBs with EPENIG plating that you have a rash of  
> soldering problems.
>
> Since you are using a WS Pb-free flux you really shouldn't be  
> having solderability issues unless your supplier is going a really  
> crappy job and your storage is terrible.  As for solder joint  
> reliability  issues when soldering to electroless platings you need  
> to make sure you are doing everything possible to make sure there  
> is minimum stress on the solder joint.  If not you are going to get  
> "Black Pad" looking brittle interface failures.
>
> -----Original Message-----
> From: TechNet [mailto:[log in to unmask]] On Behalf Of Guy Ramsey
> Sent: Monday, January 22, 2018 10:03 AM
> To: [log in to unmask]
> Subject: [TN] soldering problems
>
> So, is poor solderability like the flu . . .
> We are having a rash of it.
> We identified the EPENIG solder problem as a surface contaminate,  
> originating at the board house in the final rinse water.
> We did not get to the bottom of the ENIG problem which I am  
> thinking we own.
> Here is why.
> The "black pads" may be found on either the board or the device. I  
> have attached a photo of each, one on the board and one on the  
> device package.
>
> These are soldered with a WS Pb-free flux. We have carefully  
> developed the profile with  a long soak, to minimize the void  
> problems on LGA. Ramp, soak, and reflow are all within  
> recommendations of the solder paste manufacturer.
>
> Steve can you put these up for me. The problem does seem to be more  
> common on the DDR3 memory devices than any of the other BGA types.