Hi folks - Rockwell published a paper at the SMTA's ICSR Conference in June
on the topic of soldering PTH components with gold plated surface finishes
which addresses a number of concerns/comments in the email trail. If you
want a copy of the paper, it is in the conference proceedings. Rockwell
will also be publishing the diffusion rate of gold in SAC305 solder in a
paper titled "Dissolution Rate of Specific Elements in SAC305 Solder" at
the SMTAI Conference in October. Bottom line - gold embrittlement is a
function of solder volume/ soldering time/soldering temperature diffusion
of the gold into solder which is well understood and follows very simple
rules. The IPC JSTD-001 Handbook has a very good section on the topic as
well.

Dave Hillman
Rockwell Collins
[log in to unmask]

On Wed, Aug 1, 2018 at 12:01 PM, Stadem, Richard D <[log in to unmask]
> wrote:

> Correction: if Class 3, gold wash is required, or proof of reliability
> must be availability. (For any hi-rel application, not just Space).
>
> -----Original Message-----
> From: TechNet [mailto:[log in to unmask]] On Behalf Of Wayne Showers
> Sent: Wednesday, August 01, 2018 11:32 AM
> To: [log in to unmask]
> Subject: Re: [TN] Solder joint integrity
>
> To summarize J-STD-001, anything with a gold thickness of greater than
> flash gold (<10 micro-inches) must be gold washed in surface mount
> applications.  As for through-hole applications, the standard is higher in
> terms of thickness, but the problem of gold embrittlement must be
> assessed.  Even in surface mount applications, the issue is not in
> achieving a solder connection as hard gold (up to 18 micro-inches) can be
> soldered to fairly well, it is the concern of gold embrittlement leading to
> pre-mature joint fatigue and failure.  Hard gold is tricky as the time
> above liquidus
> The reason for this is that gold does not alloy well, rather it dissolves
> into the connection and/or creates AuSn nodules which weaken the joint
> integrity.
> I suggest contacting Leo Lambert of EPTAC ([log in to unmask]) who is one of
> the pre-eminent sources on gold embrittlement.  Also, there are several
> resources on their webpage.  Leo is the IPC Expert for much of the
> Northeast US.
>
> NOTE1: If it is for a space application, gold wash is mandated.
> NOTE2: Through-hole (bulk solder) connections typical yield a much lower
> gold percentage but, because the solder time is much shorter, the degree of
> gold diffusion is less meaning that the gold percentage left in the
> critical IMC region is much more than from SMT where the longer time above
> liquidus aids in diffusion of the gold.  This shows up most significantly
> in the feather regions of the connection where the amount of solder
> available for diffusion is the least.
>