Reversible Shape Evolution of Ge Islands on Si(001)

Abstract

The evolution of strained Ge/Si(001) islands during exposure to a Si flux was investigated by scanning tunneling microscopy. Dome islands display appreciable shape changes at Si coverages as low as 0.5 monolayer. With increasing coverage, they transform into [105]-faceted pyramids, and eventually into stepped mounds with steps parallel to the <110> directions. These morphological changes are induced by alloying and represent a complete reversal of those previously observed during Ge island growth. The results are interpreted with a simple model in which the equilibrium shape of an island mainly depends on its volume and composition.

Three-dimensional simulations of self-assembly of hut-shaped Si–Ge quantum dots

“…The experimental studies of Sutter and Lagally 8 and Tromp et al 9 on the growth of Si 1Ϫx Ge x alloys on Si͑100͒ substrates provide clear evidence for the absence of any nucleation barrier to the growth of epitaxial islands. Similar observations are reported in experiments on Ge films grown on Si͑001͒ substrates by Vailionis et al 10 and Rastelli et al 11 In general, when the Ge concentration of the films exceed 20%, growth of islands starts out via an array of shallow stepped mounds which evolve continuously to form faceted islands. More recent work of Sutter and coworkers 12 provides a clearer picture of the early stages of island evolution, where stepped mounds with atomically flat ͑100͒ terraces and single-height steps are found to act as precursors to the ͑105͒ faceted islands.…”

Si/Ge exchange mechanisms at the Ge(105) surface

“…This phenomenon was directly observed in ͕105͖ huts [35][36][37] and beautifully imaged in the case of dome-shaped islands. 34 In the latter paper, the first portion of the dome islands which was observed to be influenced by capping was the very top where highly relaxed ͕105͖ facets are present, i.e., in a region where we predict very fast Si incorporation. On more general grounds, we wish to stress how the present results, and the similar ones obtained on Si͑001͒ ͑Ref.…”

“…While any attempt to quantitatively relate our fundamental mechanisms with experiments seems difficult because of the multitude of complex and sometimes unexplored competitive processes taking place ͑typically, atomic-scale estimates of the actual Si supply through trench formation appears as par- ticularly difficult to be modeled͒, several evidences exist supporting fast Si/Ge surface exchanges at ͕105͖ facets. Beside the ones already discussed, it is worth recalling the socalled inverse shape transformation, i.e., the observed flattening of island during capping with Si, 34 as a direct consequence of the small lattice mismatch between a strongly alloyed SiGe island and the Si substrate, lowering the need of arranging the atoms in a three-dimensional structure. This phenomenon was directly observed in ͕105͖ huts [35][36][37] and beautifully imaged in the case of dome-shaped islands.…”

Shape change of SiGe islands with initial Si capping

“…This implies that the morphology change of islands with Si capping is a kinetic rather than a thermodynamic process. 9 Interestingly, when islands are capped with a 0.18-nm-thick Si layer, which is less than the thickness ͑0.32 nm͒ for a complete shape transformation for all islands from a dome to a pyramid bounded by ͕103͖ facets, some island's shapes change to a pyramid with ͕103͖ facet as denoted by A in Fig. 3͑a͒, while some islands show a shape between a dome and a pyramid with one side retaining a dome shape and the other three sides changing to ͕103͖ facets as denoted by B in Fig.…”