silver migration in components

[Robert Gliniecki]

Do you have experience dealing with silver migration?
I have a problem with a component (thermistor) which has 2 parallel
surfaces. These 2 surfaces are metallized ith fired-on silver.
A migration is occuring between the 2 silver surfaces on a perpendicular
edge.
About all I know about migration is that is occurs with low bias
voltages in the presence of moisture, but:
1. Will occur with pure (Deionized) water, or are ions necessary
for silver transport?
2. A lab analysis indicated Sodium, Chlorine, and Potassium contamination.
This probably occured during handling (not a clean room setting).
Are these ions known accelerants or promoters of migration?
3. What other conditions can accelerate migration(voltage, ion %, etc)?
4. The fixturing method (chip held by double-sided tape) may be
contributing to the problem by leaving a residue on the thermistor
chip. Any suggestions on holding down small (.040"W X 0.150"L X
0.020" thick) components for hand-soldering leads?
Help is much appreciated.