RE>>ASSY: Gold-to-Gold Soldering 10/3/96
Dave,
Good description. You're right that this is a great topic for metallurgists to discuss!
I'd like to add a few things here.
While the traditional rule of thumb is that 4 wt % of gold is the embrittlement point, there are instances where concentrations as low as ~1 w/o can create problems in solder joints. Keep in mind that the former value was obtained on bulk Izod impact samples, which led to the solder joint recommendation from years ago. I've personally
seen huge differences in solder joint integrity because of this embrittlement phenomenon with fairly low amounts of gold in them, including some joints which pretty much disintegrated when a little stress was applied to them. I seem to recall some work out of MCC many years ago (Colin McKay?) which suggested that values even lower than 1 w/o can cause problems.
The problem isn't really *how* much gold you have in the joint, but rather *where* the gold/tin intermetallics are, as you point out. Solder joint design plays an important role here.
Gold is fairly insoluble in tin and lead, I believe. I seem to recall numbers of 0.3 wt percent or so. This suggests that simply running your assembly process longer and hotter probably won't make any difference in terms of embrittlement, since it'll basically all precipitate out anyway. It's possible that doing this will redistribute the IMCs to more benign areas, though I'd guess that design would drive this if true.
The grainy appearance is most likely the intermetallic compounds orienting themselves to "poke" out of the solder. Should be very visible in microsections.
Paul Vianco from Sandia recently published an outstanding paper on this in the Journal of SMT (April 1994), "Embrittlement of Surface Mount Solder Joints by Hot-Dipped, Gold-Plated Leads". In this work, he describes the need to *triple dip* one particular gold-plated device in molten solder to eliminate this problem.
Greg Bartlett
Mercury Computer Systems
Chelmsford, MA
[log in to unmask]
--------------------------------------
Date: 10/2/96 12:54 PM
To: Greg Bartlett
From: ddhillma
Hi Greg -
Oh boy, did you ask a question that will have the TechNet hopping for
awhile!!! When you have gold finished components & gold plated PCBs
and solder them using 63/37 you run the risk of having gold
embrittlement. Gold embrittlement is one of those topics that
metallurgists and material engineers love to discuss because: (a) it
is a very well documented phenomena, (b) gold-tin intermetallics are
interesting to look at. A couple general rules of thumb to use:
a) Is there enough solder volume for the gold to diffusive into?
The answer to this question is almost always yes. The published
literature consensus is that gold embrittlement occurs at/around 4
wt.%.
b) Is the soldering operation long enough and hot enough to allow #1
to occur?
This is the question that is the problem. Many times the soldering
process is too short to allow the gold to totally diffuse into the
solder thus leaving a thin, heavy concentration of gold-tin
intermetallic where cracks will originate.
I don't have any experience with using a different alloy selection to
avoid the problems but hopefully some of the other TechNetters will
have some good suggestions. The grainy appearance of your solder
joints may be the gold-tin intermetallic particles. A good paper to
read on the subject is:
"The Effect of Gold on the Reliability of Solder Joints", Dr. Judy
Glazer, 1991 Surface Mount International Conference proceedings.
Good Luck.
Dave Hillman
Rockwell Collins
[log in to unmask]
______________________________ Reply Separator _________________________________
Subject: ASSY: Gold-to-Gold Soldering
Author: [log in to unmask] at ccmgw1
Date: 9/30/96 9:12 PM
Does anyone have experience soldering gold-plated SMT components to
gold-plated PCB's using a eutectic (Sn63/Pb37) solder? The solder
joints exhibit a dull, grainy appearance. Although I know that longer
liquidus dwell times and a maximum reflow temperature of 220 C are
typically recommended when soldering to gold, will other solder
alloys perform better (i.e,, Indium alloys)?
Thanks in advance,
Greg Kilinski
Acuson Corp.
[log in to unmask]
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to: *
* [log in to unmask] with <subject: unsubscribe> and no text. *
***************************************************************************
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to: *
* [log in to unmask] with <subject: unsubscribe> and no text. *
***************************************************************************
------------------ RFC822 Header Follows ------------------
Received: by mailgate.mc.com with SMTP;2 Oct 1996 12:53:07 -0400
Received: from firewall.mc.com (firewall [192.148.197.15]) by jericho (8.6.11/8.6.11) with SMTP id MAA08879 for <[log in to unmask]>; Wed, 2 Oct 1996 12:53:05 -0400
Received: by firewall.mc.com id AA09273
(5.65c/IDA-1.4.4 for <[log in to unmask]>); Wed, 2 Oct 1996 12:53:00 -0400
Received: from ipc.org(168.113.24.64) by firewall via smap (V1.3)
id sma009262; Wed Oct 2 12:52:15 1996
Received: from ipc.org by simon.ipc.org via SMTP (940816.SGI.8.6.9/940406.SGI)
id LAA23409; Wed, 2 Oct 1996 11:31:22 -0700
Resent-Date: Wed, 2 Oct 1996 11:31:22 -0700
Received: by ipc.org (Smail3.1.28.1 #2)
id m0v8TNz-0000QbC; Wed, 2 Oct 96 10:38 CDT
Resent-Sender: [log in to unmask]
Old-Return-Path: <[log in to unmask]>
Date: Tue, 01 Oct 96 08:18:18 cst
From: "ddhillma" <[log in to unmask]>
Encoding: 70 Text
Message-Id: <[log in to unmask]>
To: [log in to unmask]
Cc: [log in to unmask]
Subject: Re: ASSY: Gold-to-Gold Soldering
Resent-Message-Id: <"Xn65x1.0.R4F.FoeKo"@ipc>
Resent-From: [log in to unmask]
X-Mailing-List: <[log in to unmask]> archive/latest/6628
X-Loop: [log in to unmask]
Precedence: list
Resent-Sender: [log in to unmask]
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to: *
* [log in to unmask] with <subject: unsubscribe> and no text. *
***************************************************************************
|