Marcello

Quite a while ago we had a discussion about the effects of the cooling rate
onto the reliability of solder joints and you asked me for further
comments. Sorry, I didn't want you to wait such a long time. Somehow time
is just flying.

In tin- Lead solder the tin-lead phases make the largest contribution to
the deformation of the material. The Lead phases are much more difficult to
deform. These tin-lead phases show primarily two deformation mechanisms:

- Grainboundary sliding (GBS) at elevated temperatures and slow deformation
rates
- Dislocation climb (DC) at low temperatures and high deformation rates

If a solder joint in SMT does solidify the material has no time to form
small crystals. Large tin-lead domains will form with small lead-tin phases
within. In fact I have seen many solder joints consisting of 2-3 domains.

If these solder joints are subject to deformation GBS can not occur, since
virtually no grain boundaries are present. Consequently the material is
enforced to deform in the DC regime. A consequence of DC deformation is the
formation of micro-voids due to the accumulation of dislocations at the
grain boundaries as well as at inhomogenities in a domain. The latter will
cause the formation of new grain boundaries thus recrystallisation of the
tin-lead domain in the deformed zone the first will cause a reduction of
the load bearing area. This means a concentration of the deformation in
this area and further accumulation of dislocations. Naturally also GBS has
his damaging mechanism. You may see this if you build a puzzle of unevenly
formed pieces and try to slide these pieces above each other. You will see,
that they don't fit at all. Voids are formed. In creep deformation in the
GBS regime these voids are filled by diffusion in the solid body from areas
where the crystals have to endure pressure to the areas where there is
tensile stress. However, diffusion is a time dependent process. If the GBS
is running too fast, the voids are only filled partially again reducing the
load bearing area. We believe that this mechanisms are present if a crack
grows only through the solder. Once the crack reaches the solder-component
interphase the whole thing becomes very complicated since there you have
the intermetallics reaching into the solder and you don't have the original
solder alloy. Due to the formation of the tin-copper intermetallic layers
the solder at the interphase contains more Lead. But not much work has been
done so far in this field.


Best regards

Guenter