My department and I have been working with immersion silver on and off for about four years and we have compiled quite a bit of data for the Alpha and MacDermid immersion silver chemistries. We have not encountered any assembly or storage related issues that could not be easily overcome. But we have found that the long-term reliability of our solder joints (Sn/Pb) assembled to immersion silver can be dramatically different than the joints assembled to ENIG or Cu OSP and the reliability can very dramatically from test to test (whereas the Cu OSP and ENIG give consistent results). The silver seems to give us an increased failure distribution time which is the result of very early failures and very late failures (vs. Cu OSP). Has anyone else encountered such a problem? It seems to be related to silver thickness (which affects the volume of organic co-deposit), but we cannot conclusively prove this.