The authoritative public forum
  for Metal Finishing since 1989
-----

Effect of Heat on Gold/Nickel Plated Kovar

I am have a discussion with other engineers with the affect of heat on a nickel / gold plated TO-5 header. The base metal of the header is either kovar or Alloy 42 with kovar leads. It is electroplated in a barrel of about 2,000 headers at a time. the plating is 50 to 100 micro inches of nickel followed by 50 micro inches of high purity gold.

In our manufacturing process we present submit these headers to a temperature of 450 degree C for 3 minutes (in a graphite boat). I believe that the heat is causing the gold to diffuse into the nickel, therefore causing problems in thermocompression wire bonding.

I need to find somewhere this or something similar has been documented.

Do you have any suggestions?

Thank you,

Steve Hall
- Carrollton, Texas
1998

---

1998

Steve,

Is the heating done in vacuum? It is well known that when heating Cr/Au PVD metallization in air the Cr will diffuse through the Au to the Au surface where it will oxidize and give bonding problems. This can take place at temperatures as low as 200 deg. C. The diffusion does not occur in vacuum where there is no oxygen to react with the Cr. I imagine that nickel will do the same.

Donald M. Mattox
Society of Vacuum Coaters
Albuquerque, New Mexico

I would like some clarification to the above answer. How would the lack of oxygen prevent the Cr from diffusing to the surface? I understand that you would not form the oxide in a vacuum. However, I do not understand how the presence of vacuum limits the diffusion to the surface.

John Drosdak
- Corning, New York
2001

The basic purpose of the nickel coating is to provide high temperature barrier to gold from diffusing in and copper to diffuse out toward gold under practical ambient conditions. So I would investigate nickel thickness and uniformity first and then check the oven/boat for contamination. Thick nickel may help. Graphite should be reacting with oxygen in air and wear out. A neutral environment, nitrogen or forming gas could be better.

Mandar Sunthankar
- Fort Collins, Colorado
2001

---

I've seen this issue in the past at temperatures near 450C (in our case it was on Ni/Au plated flexible circuitry). What happens is that nickel diffuses to the surface and oxidizes (you can look up the diffusion rate for nickel in Au in a CRC to convince yourself). A thin film of nickel oxide forms on the surface and creates difficulty with gold ball bonding. We fixed the problem by cleaning the circuits in a 20% sulfuric acid solution - however we found that an argon plasma would also do the job. The best way to confirm the results is with depth profiling Auger.

Good luck.

Randy Schueller
- Austin, Texas
2001

I have seen this problem before. You should not heat the part in an oxygen environment because you will create an oxide layer on the diffused nickel. You should heat the part in Nitrogen.

Nancie Barker
- Morgan Hill, California
May 28, 2008