John and technetters,
I'll try to address this question from my experience within the high
temperature environment.
Electro-Plated Nickel (Ni)
The first circuit finish that we evaluated was electro-plated
Nickel. With a high melt point and showing no significant diffusion
into any of the readily available medium temperature (melt point above
220oC and below 350oC) solders it seemed at first glance an ideal
solution. Nickel was common to most all PCB fabricators as it was
commonly plated above the copper and below the Gold on edge contacts,
so suppliers could supply the required finish with little or no
problem. There was however a fly in the soup. Nickel by itself is very
difficult to solder without the use of very corrosive fluxes. The
problems with removing the flux residue led to the next solution.
Gold over Nickel (Au/Ni)
If a PCB fabricator could plate Gold over Nickel on the circuits
just like they did with edge connector, then the OEM could solder to
the circuit. The gold would dissolve into the solder and the real
solder joint would be made of Nickel. Problem solved or so it seemed.
As other developments allowed, the tools were taken to higher and
higher temperatures for longer periods of time and problems with the
solder joint began to appear. The most commonly used solder at this
point in time was "HMP", an alloy of about 94% Lead 1% Silver and 5%
Tin with a melt point of about 588oF (309oC). The "standard" circuit
finish at this time was a minimum of 50 uinches of Gold over a minimum
of 300 minches of Nickel. The primary problem that had appeared was
cracks in the solder joint. This it turned out was caused by Gold
embrittlement. The combination of a high percentage of Lead in the
solder, the amount of Gold and the increased time at elevated
temperature had brought the problem to light. The first solution was
to decrease the amount of Gold on the Nickel to 20 uinches ± 4
minches. The Gold embrittlement of a high Lead solder is a non-linear
function, by reducing the Gold to less than 1/2 of the previous amount
the embrittlement was reduced by a factor of 4. The Gold thickness
could not practically be reduced further because of the reduced
storage time allowed for an unsoldered PCB. The thinner the Gold the
faster the Nickel would passivate. 20 uinches of Gold allowed a useful
shelf life of 3 months or so.
Tin/Nickel (Sn/Ni)
Tin-Nickel plating was recommended as a substitute for the
Nickel/Gold finish. This finish is still occasionally incorrectly
referred to a the "Santa Clara" process. The process patented by Santa
Clara circuits did involve Tin-Nickel plate but also called for
selective plate of "standard" Sn/Pb at the solder points. The
Tin-Nickel seemed to be a good solution, easier to solder than pure
Nickel and less expensive than Gold plating the Nickel. Because the
Tin and Nickel are plated at the same time in the same plating bath
and due to the characteristics of the two metals they are plated as a
combination of intermetallic compounds. The most common of these
intermettalics is NiSn, however with time and temperature NiSn
transforms into NiSn22. This new molecule occupied a greater volume
than did its components. This induced extreme internal stresses and
the plating would micro crack and a fine conducting dust would be free
to spread over the board. This problem was of no concern and was not
noticed at temperatures used with FR-4 materials. It takes about 100
years for the transformation to go on to completion at 85oC, however
complete transformation will occur in a little over 200 hours at
200oC. The longer the part was at elevated temperature the faster the
reaction occurred. Once this phenomena was identified the use of the
co-deposit of Tin-Nickel fell into disfavor.
Nickel/Palladium/Gold (Ni/Pd/Au)
There had to be a better way. Enter Nickel / Palladium / Gold
finish first developed by Bell Labs as part of their effort to reduce
the cost of precious metals in their equipment. Their work showed that
Palladium with as little as 3 to 5 uinches of Gold on top made an even
better contact for switches and relays than 40 uinches of Gold over
Nickel. Secondly Palladium was easy to solder and cost about 1/4 as
much as Gold. The solution was quite simple, plate Nickel over the
copper to act as a corrosion and diffusion barrier, plate Palladium
over the Nickel as the surface to take the solder and plate the
Palladium with Gold (soft) to further improve the overall
solderability. We now have the perfect (well almost) finish for use
with high temperature solders, there is so little Gold that
embrittlement of Lead solders ceases to be a real problem and the high
Sn-based solders work well with this finish as well as some of the
more "exotic" solders.
D.Drake
SSDS Houston
______________________________ Reply Separator _________________________________
Subject: RE: Assy: Au/Pd/Ni/Cu PCB Finish
Author: John Guy <[log in to unmask]> at Internet-Mail
Date: 6/12/97 8:31 AM
Bill and Technetters,
I'm sure I am showing my ignorance when I ask this, but if I don't, how
else can I learn. My questions is this. What is the advantage of having the
palladium layer in there? Does it have something to do with the plating
operations?
Best Regards
John Guy
American Competitiveness Institute (EMPF)
714 N. Senate Ave.
Indianapolis, IN 46202-3112
Tel: (317) 655-EMPF x130
fax: (317) 655-3699
e-mail: [log in to unmask]
----------
From: Barthel, Bill[SMTP:[log in to unmask]]
Sent: Thursday, June 12, 1997 7:45 AM
To: 'TechNet'
Subject: Assy: Au/Pd/Ni/Cu PCB Finish
Does anyone have experience with assembly and long term reliability of
surface mount and through hole solder joints made on PCBs coated with gold
over palladium over nickel over copper?
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To subscribe/unsubscribe send a message <to: [log in to unmask]> *
* with <subject: subscribe/unsubscribe> and no text in the body. *
***************************************************************************
* If you are having a problem with the IPC TechNet forum please contact *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask] *
***************************************************************************
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To subscribe/unsubscribe send a message <to: [log in to unmask]> *
* with <subject: subscribe/unsubscribe> and no text in the body. *
***************************************************************************
* If you are having a problem with the IPC TechNet forum please contact *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask] *
***************************************************************************
|
