recently I met one very interesting issue with my board and connector. substrate of the board is copper with immersion silver, pad size is about 3x1mm. we print lead-free paste on the pad, then put the connector on the paste. The board was reflowed under a profile wiht a 65 sec of dwell time, some leads of the connector (at a ppm level)was pull out with the solder. all the solder was totally sticked to the lead and pell off from the pad, nothing was left on the pad. under the micoscope, it was observed the solder was reflowed and wet well. for the good joint, the solder will stick to the pad if we try to pull out the lead from the joint. anyone have ever meet the issue like this, any idea about this issue, what is the possible reason for this issue? waiting for your reply!!
First as a nit, your �substrate� is probably NOT copper. It�s probably FR4 fiberglass/epoxy.
Substrate. Base Material. A supporting insulating material upon which parts, components, and elements are attached.
Second, enough of that, on to your issue � comments are: * �Dwell� [time above liquidous???] is none too long. We�re not experienced with no-lead, but we�d expect liquidous +35*C to be 5 to 10 sec. * That your solder is not sticking to the pad and is globbed on the lead indicates  the lead is solderable and the pad is not. And/or  the lead is much hotter than the pad.
In  above, you should be able to test the solderability of the pad. It�s possible that the imm Ag was plated over marginally unsolderable copper and over time, it became less solderable.
In  above, assess this by measuring the temperatures of your reflow temperature recipe at the discrepant pads.