X-ray topography of the coherency breakdown in GexSi1−x/Si(100)

Abstract

An x-ray topography study is presented of the coherency breakdown in GexSi1−x/Si(100) strained epilayers. Finite dislocation densities (in excess of 103 cm−2) are observed at compositions in the range 12–13 at. % Ge for an epilayer thickness of h≊180 nm. Above 13 at. % Ge the dislocation density starts to change rapidly and this composition is identified as critical for h≊180 nm, a thickness which is almost a factor of 4 lower than the accepted ‘‘critical’’ thickness for this lattice mismatch. The result suggests that in low-mismatched GexSi1−x alloys the dislocation density will increase continuously at the ‘‘critical’’ thickness, as opposed to exhibiting a sharp onset. The implications of these results to the various models of the critical thickness transition are discussed.

Does In form In-rich clusters in InGaN quantum wells?

“…Electron microscopy of the InGaN quantum wells we have studied in this paper reveals no misfit dislocations. We are aware that the measured critical thickness for the introduction of misfit dislocations depends on the resolution of the experimental technique used to detect the dislocations [33] and that electron microscopy, because of the limited volume of specimen sampled, may over-estimate the critical thickness. However, electron microscopy indicates that, at least locally, our InGaN quantum wells are fully strained, and this is confirmed by our X-ray diffraction measurements.…”

Nucleation of semicircular misfit dislocation loops from the epitaxial surface of strained-layer heterostructures

“…These sources can then provide misfit dislocation formation if the requirements of the critical thickness and the energy barrier are satisfied. Matthews and co-workers 6 and Eaglesham et al 22 pointed out that, for a given temperature, there exists a critical misfit strain ⑀ c , below which the nucleation cannot be thermally activated and the critical thickness will be infinity. In Fig.…”

Effect of Si doping on the relaxation mechanism of InGaAs on GaAs

“…10,11 It would appear that the stress required to initiate cross slip is not a function of Si concentration. We may expect that the modified Mat-thews and Blakeslee model will break down for the B(g) misfit dislocations beyond a Si concentration of approximately 6ϫ10 18 cm Ϫ3 as, at this point, the stress required for cross slip and nucleation of the slow A(g) set will equal the stress necessary to drive the threading segment across the wafer in the ͓110͔ direction.…”