Small amounts of Ge atoms are deposited on Si͑111͒-7 ϫ 7 surfaces at room temperature ͑RT͒ and at 100°C to clarify the initial adsorption sites using scanning tunneling microscopy. At RT Ge atoms are adsorbed at high coordination B 2 sites around the rest atoms, as predicted by Cho and Kaxiras ͓Surf. Sci. 396, L261 ͑1998͔͒. On one hand, at 100°C Ge atoms are adsorbed on corner adatom sites. With increasing Ge coverage the corner sites are gradually occupied, followed by Ge adsorption at center adatom sites, resulting in Ge cluster growth with a size of the half unit cell. © 2006 American Institute of Physics. ͓DOI: 10.1063/1.2198109͔The stable periodic structure of a Si͑111͒-7 ϫ 7 surface with a comparatively large size unit cell ͑UC͒ offers a good template for fabricating self-assembled nanostructures. [1][2][3][4]6 For example, Si magic clusters and self-assembled metal nanostructures confined in a UC over the Si͑111͒-7 ϫ 7 surface have been extensively studied. [3][4][5][6][7][8][9][10][11][12][13][14] In addition, there are many reports on the initial Ge growth in the UC of Si͑111͒-7 ϫ 7 accompanied by the formation of clusters with various shapes and sizes at submonolayer coverages, 3,4,[13][14][15][16][17][18][19][20] followed by Stranski-Krastanov ͑SK͒ growth mode with increasing Ge coverage owing to Si-Ge lattice mismatch.At present, there is a considerable controversy regarding the initial Ge adsorption site and the bonding coordination over the Si͑111͒-7 ϫ 7 surface. Dev et al. have performed x-ray standing wave measurements at submonolayer Ge coverage and proposed that Ge adatoms directly bond to Si adatoms and rest atoms. 12 However, using first principles density-functional theory ͑DFT͒ calculations, Cho and Kaxiras suggested that a bridge ͑B 2 ͒ site among the high coordination sites has the minimum energy for Ge adsorption, even compared to those of Ge adatoms bounded with dangling bond of adatoms and rest atoms. 21 On the other hand, Wang et al. 20 explored the Ge adsorption over Si adatom sites in detail as well as sites other than the B 2 site. They concluded that Ge atoms are replaced with Si adatoms, and the kickedout Si atoms migrate to near B 2 sites within a basin on a half unit cell ͑HUC͒: This is in agreement with earlier reports. 3,4,15,20 In this letter we report on the initial stages of Ge adsorption on Si͑111͒-7 ϫ 7 at room temperature ͑RT͒ and at 100°C, which are observed by scanning tunneling microscopy ͑STM͒. The results show that Ge atoms are adsorbed over B 2 sites at RT, while they are adsorbed at corner adatom sites at 100°C by being presumably replaced with the Si adatoms, preferably over a faulted half unit cell ͑FHUC͒.Experiments were performed using a homemade ultrahigh vacuum ͑UHV͒ STM having a base pressure Ͻ2 ϫ 10 −11 Torr, equipped with a field emission microscope 22 ͑FEM͒ to evaluate STM tips, and with a resistive-type Ge evaporation source. N-type Sb-doped Si͑111͒ substrates were used ͑thickness= 375 m, = 0.02 ⍀ cm͒. Clean Si͑111͒-7 ϫ 7 surfaces were prepared by the ...