Answers are grounded in this paper and cite their sources.
We have quantified the strain and the defects present in tensile-strained Si layers grown by reduced pressure–chemical vapour deposition on polished Si 0.6 Ge 0.4 and Si 0.5 Ge 0.5 virtual substrates (VS), as a function of their thickness. UV-Raman spectroscopy has shown that the tensile strain in the sSi layers decreased very slightly as the sSi layer thickness increased. Mean values, 2.58 GPa for sSi on Si 0.60 Ge 0.40 and 3.15 GPa for sSi on Si 0.52 Ge 0.48 , are very close to those expected for fully pseudomorphic sSi layers. We have also studied, thanks to Secco etching, the defects present in those sSi layers. The threading dislocation density (TDD) slightly increases from 1 up to 3 × 10 5 cm −2 as the sSi thickness increases. It is, however, rather close to the TDD associated with the SiGe VS underneath, i.e. 1.85 × 10 5 cm −2 . Meanwhile, the line defects (i.e. stacking faults and/or misfit dislocations) linear density (LLD) increased from 0 up to 800 cm −1 (from 0 up to 300 cm −1 ) as the sSi layer thickness on Si 0.60 Ge 0.40 (on Si 0.52 Ge 0.48 ) increased from 5.6 up to 28.4 nm (from 4.7 up to 23.7 nm). Differences might be due to the long-range, small amplitude surface cross-hatch remaining on Si 0.60 Ge 0.40 VS after chemical mechanical polishing. By contrast, the Si 0.52 Ge 0.48 VS surfaces were featureless. Were we to adopt as some sort of critical thickness those for which the LDD exceeds 50 cm −1 , we would find values around 9 nm irrespectively of the Ge content.
No takes yet. Share an insight, caveat, or question.
Hartmann et al. (2007) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: