There is an excellent article on surface finishes by Paul Vianco of
Sandia National Laboratories in Albuquerque. It is in the October 1998
issue of the PCIF Journal Circuit World published here in the UK.

As I understand it, the ordinary ENIG (Electroless Ni/Immersion Au)
finish is more suitable for soldering and Aluminium wire bonding
whereas there are issues of joint embrittlement when soldering to
Palladium. This is because it dissolves more slowly and does not
disperse as rapidly as gold leaving it concentrated at the joint
interface.(Whereas 3-4% Au content in solder is the recognised limit,
it is only 1-2% with palladium). The Aluminium wire bonds directly to
the nickel surface so a thin gold coating is better to avoid
interference with the bond strength.

The main benefit of palladium appears to be that it is a better finish
for gold wire bonding provided it is 0.25-0.6 microns thick. To
achieve this, an electroless rather than an immersion palladium
process is required. Immersion Gold is then applied over palladium to
protect the surface from oxidation. The gold wire is bonded to the
palladium, not the gold.

Gold wire can be bonded to ENIG provided the gold thickness is
sufficient. In this case it bonds to the gold, not the nickel. As the
gold can be between 0.075 - 0.15 microns thick, there is much less of
it to cause solder joint embrittlement. Thicker gold coatings require
an electroless gold process which is better for gold wire bonding.

All of the above mean that the palladium finish is more expensive the
ENIG but can't say by how much.

I hope I have not misinterpreted Paul Vianco's detailed assessment
which is well worth studying.

Regards
Paul Gould
Teknacron Circuits Ltd
UK
www.teknacron.com

-----Original Message-----
From: Au Po Lam, Benny <[log in to unmask]>
To: [log in to unmask] <[log in to unmask]>
Date: 29 January 1999 07:25
Subject: [TN] Palladium underplate for Cu-Ni-Au system


>Hi technetters,
>
>Does anybody know the purpose of a palladium layer between the Ni and
Au for
>wire bonding PCB? Recently, I've got a request to bond gold wire
>(thermosonic) on PI flex which is plated with electroless Ni and
immersion
>Au. On one sample, the gold thickness is reduced and compensated by a
>electroless palladium layer above Ni; while the other just is the
normal
>electroless Ni and immersion Au. The phosphorus contect is around
6-8%.
>
>I would appreciate if anyone can advise me on the palladium function,
cost
>implication (cheaper by reducing the gold?), and bondability of this
kind of
>metalization.
>
>Regards, Benny.
>ASM Assembly Automation
>Email :[log in to unmask]

################################################################
TechNet E-Mail Forum 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's web site (http://www.ipc.org) "On-Line Services" section for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################