Authored By:
Phil Isaacs, Jing Zhang & Terry Munson
IBM Corporation & Foresite, Inc.
Summary
Dendrites, Electrochemical Migration (ECM) and parasitic leakage, are usually caused by process related contamination. For example, excess flux, poor handling, extraneous solder, fibers, to name a few. One does not normally relate these fails with environmental causes. However, creep corrosion is a mechanism by which electronic products fail in application, primarily related to sulfur pollution present in the air.
The sulfur reacts with exposed silver, and to a lesser extent, exposed copper. This paper will explore various
aspects of the creep corrosion chemical reaction:
- What is driving the creep corrosion reaction?
- Why is drying the product a necessary precursor to obtaining creep corrosion in tests?
- Test methods with Flowers of Sulfur, FoS, and sulfur rich clay.
- Discussion of creep corrosion related field fails.
- When does creep corrosion become ECM.
- Sources of sulfur containing pollution.
- Methods to take to avoid creep corrosion.
While there are places with sulfur containing pollution, creep corrosion will be a factor which will impact reliability. Creep corrosion will need to be understood and handled.
Conclusions
The preconditioning environment prior to subjecting the parts to a sulfur containing environment made a significant difference in the amount of creep corrosion which grew on the samples.
- All the samples preconditioned with an 40C, 90% relative humidity environment had far less creep corrosion than their counterparts in the test which were dried prior to exposure.
- The QFN mounting pads tended to have less creep
corrosion than the corresponding PTHs on the same test vehicles. - Where the hydrophobic nature of elemental sulfur
may explain the lack of creep corrosion that occurred on the samples preconditioned with temperature and humidity, it does not explain why the samples that were submitted to MFG, hydrogen sulfide and sulfur dioxide gases, also had far less creep corrosion than the dried counterparts in the experiment. - The parts preconditioned with 40C and 90% R.H. may have accelerated the formation of a copper oxide layer, forming a “passivation” layer, retarding also the corrosion in the MFG test.
- There are various approaches to prevention of creep corrosion. The most effective methods preclude the contact between the atmospheric sulfur and the printed circuit board assemblies.
- The results of all three test methods correlate well, regardless of the form of sulfur containing pollution.
- In the FoS test, the sodium hypochlorite appears to have participated in the reaction on the samples prepped with 40C and 90% R.H., making cupric chloride.
Initially Published in the SMTA Proceedings
|
