Gold Embrittlement

Intermetallic growth is a diffusion process, the thickness of the intermetallic is given by: z^2 = (Do)(t)exp[-Q/RT] Where Do (m2 s-1) is the diffusion coefficient, T(�K) is the temperature, Q (J mol-1) is the activation energy and R (8.314 J mol-1 �K-1) is the gas constant. From previous work at CALCE, recommended values for the activation energy for 60Sn40Pb solder is 78 kJ and the diffusion coefficient was empirically calibrated as 1.5 x 10-5 m2 s-1 . If you plug and chug, you will find that at 80*C (353*K), Time||IMC Thicknss micron 10 sec||0.02 1000 sec (~15 min)||0.2 100000 sec (~1 day)||2 10000000 sec (~100 days)||20

While CALCE reports are usually numbingly expensive, what do you loose?

Cemal Basaran has talked on SMTnet about measuring the thickness of IMC in good / bad connections. So that may be someone else to check with: Cemal Basaran, Dr Associate Professor SUNY Buffalo Electronic Packaging Laboratory

Work: 716-645-2114 X2429 Fax: 716-645-3773 E-Mail: cjb@eng.buffalo.edu

243 Ketler Hall North Campus SUNY at Buffalo Buffalo, NY 14215

reply »

To maybe simplify, contact David Sbiroli(Apps. Engineer) @ Indium corporation. I have worked with Dave on this issue before. He can most likely get you the info you need.

1-800-4 INDIUM

PS: Dave F. : My plasma arc accumulator has unusual carbon scoring on and around the dylithium chamber plenum. Any ideas ?

reply »

There are studies out there that looked at gold embrittlement. The overall consensus is to not exceed 5% by weight. As far as the conductivity question, as themo/mechanical creep fatigue sets in, your conductivity will be affected as the joint is separating until 100% fracture. Thermal cyles, of course, will augment the difusion process (and embrittlement) as Dave (so eloquently) described. I gotta run, oddly, the gamma ray alarm just sounded and I'm getting very ANGRY right now!

reply »