An initial phase of Ge hut array formation at low temperature on Si(001)

Abstract

We report a direct STM observation of Ge hut array nucleation on the Si(001) surface during ultrahigh vacuum molecular-beam epitaxy at 360C. Nuclei of pyramids and wedges have been observed on the wetting layer MxN patches starting from the coverage of about 5.1 \r{A} (~3.6 ML). Further development of hut arrays consists in simultaneous growth of the formerly appeared clusters and nucleation of new ones resulting in gradual rise of hut number density with increasing surface coverage. Huts nucleate reconstructing the patch surface from the usual c(4x2) or p(2x2) structure to one of two recently described formations composed by epitaxially oriented Ge dimer pairs and chains of four dimers.Comment: Extended discussion; to appear in J. Appl. Phys., 201

On atomic structure of Ge huts growing on the Ge/Si(001) wetting layer

“…1,4,10,15,28 Our STM data demonstrate these two cluster forms to grow from different types of nuclei which have different structures and symmetries: 15,18 Fig. 1 presents STM images of these nuclei and models of arrangement of their atoms on WL.…”

“…7) when the clusters are small enough and the distances between them are large in comparison with their dimensions. [15][16][17][18] In these conditions, if a flux of Ge atoms arriving on the surface is sufficient to feed all the growing huts and a competition between huts for Ge is actually absent, all pyramids, both stable and metastable, obtain enough material to complete their facets and grow remaining in the array. Pyramids of 6 ML high are stable but they form at greater coverages, from 6 to 8Å, when all huts become bigger, more actively consume Ge and gaps between then become smaller.…”

“…1). Unfortunately, no satisfactory explanation of this phenomenon has been proposed thus far which would describe a driving force of this strange behaviour of Ge dimers which group in two different formations, 18,19 instead of a single one, on tops of WL patches after the latter have reached their maximum thickness. 23 We have demonstrated that both pyramids and wedges grow in height conserving the width and the atomic structure of their apexes (the topmost (001) terraces).…”

“…Our recent experimental explorations 6,[15][16][17][18][19]21 carried out on atomic level by high-resolution STM have demonstrated fine details of hut nucleation and its evolution during the growth which however have not been sufficiently interpreted thus far and presented in terms of structural schematics and drawings which would be helpful for further theoretical calculations and numerical simulations. For example, we have discovered the phenomenon of simultaneous appearance of two types of nuclei 15,19 which are composed by 16 dimers and different only in symmetry-a separate nucleus for each species of huts-on tops of Ge WL M × N patches 22 of 4-monolayer 18 (ML) height (Fig.…”

“…However, only a few of these works studied this issue on atomic level. [15][16][17][18][19] Investigations carried out on atomic-level were mainly devoted to the structure of the {105} facets, for which an initially proposed simple model based on paired dimers 1 (the so called PD model) was eventually replaced by a model considering step rebonding as a source of the facet stability 20 (the latter one is usually referred to as the RS model), rather than to nucleation of the clusters or their in-height or longitudinal growth.…”