Question about dislocation nucleation sites in strained silicon-on-insulator

Electronic active device is built on the strained silicon-on-insulator (sSOI), e.g. strained Si layer on oxide, which in turn is bonded on bulk silicon wafer. Because no misfit dislocation can exist in strained silicon layer any more, will the dislocation be generated during later processing and operation? If there are still lots of dislocations in the strained silicon layer, where do they come from? Is there any experimental work to discover the dislocation nucleation sites? I guess they will nucleate from the triple junctions of gate-sSOI-cap, because the stress is singular in the triple junction. But I am not sure. So I want to know something about the experimental observations.

Thanks in advance if anybody here can help me.