As evident, within the experimentally significant volume range, dots are always more stable than wires. This is due to their lower surface area per unit volume (about Rabusertib 40% less) compared to the wires (Table  1). The measured surface to volume ratios match well with those expected for ideal 113 wires and islands. The analysis, thus, confirms that the wires are metastable structures which are formed solely due to the

presence of the preexisting polishing-induced defects. In the presence of tensile epitaxial strain induced by Si deposition, the wires thus evolve into the stable dot shape which allows a more efficient strain relaxation. Conclusions In summary, we have described the quite complex mesoscale structure of Ge(001) substrates cleaned by sputtering/annealing treatments, indentifying the sputtering-induced defects and distinguishing them from polishing-induced intrinsic defects. By positively exploiting the polishing-induced defects of standard-quality commercial Ge(001) wafers, micrometer-length Ge wires can be grown without introducing any metal catalyst. The shape of the wires can be tailored by the epitaxial strain induced by Y27632 subsequent Si deposition, determining a progressive transformation of the wires in SiGe faceted quantum dots. We remark that the spatial distribution of the wires (i.e., direction, spatial ordering, etc.), and therefore of the dots formed by Si overgrowth, are dictated

by the characteristics of the polishing-induced trenches. As a future perspective, controlling the polishing feature will therefore enhance the spatial ML323 clinical trial ordering of nanostructures. Acknowledgements The authors acknowledge the support of Dr. H. Diao, Dr. J. Riches, and Dr. L. Rintoul from the Central Analytical Research Facility (CARF) at QUT for FIB, TEM, and Raman characterization, respectively. LP acknowledges the support from the ETH Zurich Postdoctoral Fellowship Program and the Marie Curie Actions for People COFUND Program. NM and MN acknowledge the financial support of the Australian Research

Council through the Discovery Project DP13010212. Electronic supplementary stiripentol material Additional file 1: Surface morphology obtained by different cleaning treatments. Comparison of large-scale surface morphology obtained by different cleaning procedures: (a) 4 cycles Ar sputtering (830 V, 20 min, 2 × 10-7 mbar Ar) and subsequent annealing at 830°C for 20 min. (b) 8 cycles Ar sputtering (830 V, 20 min, 2 × 10-7 mbar Ar) and subsequent annealing at 830°C for 20 min. (c) Ex situ chemical passivation followed by an in situ heating procedure. A GeOx passivation layer is chemically grown ex situ by a wet treatment consisting of a HCl/H2O 36:100 bath and subsequent H2O2/H2O 7:100 bath to strip/reform a GeOx passivation layer. The samples are then outgassed in situ at 230°C for 1 h, flash annealed at 760°C for 60 s to remove GeOx, and slowly cooled from 600°C to room temperature. (PDF 451 KB) References 1.