Gaby,

Also noticed: we learned never to belittle the AuSn-formations because of a thin Au-plating. Some thought that a poor one-micron Au will dissolve in a massive Sn/Pd joint and 'disappear' as a harmless minority. Was not always true, because we found in some cases, that the Au did not at all dissolve and spread equally, instead we got high concentrations of Au near the interface between lead and Sn/Pd and also dark, brittle intermetallics that caused cracked regions. Therefore, degolding had to be done before soldering. As this is not a very economic way, most parts are ordered in pretinned fashion today. So, SPACE, MIL or Commercial, we try always to avoid goldplated terminals today (in earlier days we loved golden terminals).

Ingemar

PS. I have no explanation for the above behaviour of gold that did not spread equally, maybe impurities in the solder...


--------------------------------------------------------------------------

There are many practical problems in discussing solder compounds: 1) when
does one talk about "equilibrium" compositions/phases when one has
nonuniform thermal processing (heat/cool), 2) interface metals not fully
dissolved, 3) gaseous and liquid (flux) impurities, and 4) reliable aging
tests.

That said, here is my comments on tin-gold:
The issue with using gold is the occurance of the crystal structure of the
AuSn4 phase depending on the solder mix, about 220C melting/liquidus/phase
change point. Above a critial value (approximately 5% Au for Pb37-Sn63 wt%
or ~10% Au for pure Sn), the crystal structure becomes large grained.  The
the poor interfacial properties between AuSn4 crystals tend to dominate the
mechanical strength, leading to embrittlement.  (For the record, at lower
percentages of Au, you still have AuSn4 crystals, only that they are
fine-grained and are dispersed in the Pb-Sn mix.  In fact, they can increase
the quality of the joint by "pinning" the Pb-Sn grains from moving without
much plastic flow).  So soldering to pure Au with Pb-Sn (or pure Sn) is
going to get you into trouble.  Track down some of the past work on TAB
bonding to Au-bumped Si substrates (sorry, I don't have any references to
start you on).

Still I'm confused about the original memo (below in a mail response).  If
you are soldering to gold, you are starting from the worst alloy in using
AuSn4, the most brittle form of Au-Sn.  If you want to solder to gold, use a
gold-dominated solder (e.g. Au5Sn).  But these are in the 400C+ melt region.
These are typically used as initial/tack-down solder step or in packaging?
If so, then they can go through a secondary soldering (Pd-Sn) step at 200C
without
noticeable damage (assuming the Pd-Sn is not applied to the Au area).

Any comments from the peanut gallery?
Carey

##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in
the body:
To subscribe:   SUBSCRIBE TECHNET <your full name>
To unsubscribe:   SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################