One may use ECCI or TEM (smaller field of view than ECCI) to determine an average density of the dislocations for a set of samples with different densities (variation of lattice mismatch, growth process, etc.)and compare it with CL or PL spectra of the the same set of samples.

If CL imaging is possible, then panchromatic or even monochromatic CL images can be used to count the dislocations at the film surface. This was working fine for me for GaAsP samples with different substrates, i.e. different lattice mismatches. When the TDD becomes to high, counting of the dislocations becomes more difficult.

Another but more sophisticated approach would be to perform correlative TEM-CL or ECCI-CL measurements.

I hope this will help a bit.

Stefan,

how do you intend to distinguish the TDDs contribution to YL from the carbon contribution?