Applied Surface Science 6(I/61 ( 1992t 107-11 I North-Holland ap#~-=d surface Science Growth of Si on Au deposited Si(lll) surfaces studied by UHV-REM H. Minoda, Y. Tanishiro, N. Yamamoto and K. Yagi Ph~ics Department, Tokyo ltlstiotte Of TechnologY. Oh.oktz~ao!a, Megltro. Tokyo 152. ]anu~l Received 211 November 1991; accepted for publication 30 January 1092 Growth of Si on Au-deposited St(Ill 15 × 2 surfaces ~,as studied by ultra-high-vacuum reflection eleclron microscopy, Above 40ICC RHEED sho~ed the 5 x 2 patlcrn during the Si deposition indicaling Ihal the AU deposits stayed at the top surface during depositinn, Suppression of t~o-dlmensional nucleation of Si to some extent was noted, preferential nuelealion of 2D island~; at orientalioual domain boundaries of the 5 × 2 structure ~ , obse~ed. 1. Introduction A recent interesting problem in studies of epi- taxial growth is surfactant-medlat•d epitaxy [1 3]. In this case growth is on a surface which has been pro-covered by a so-called surfactant material. During the deposition of the other materials the surfactant always stays at the top surface, which indicates that the growth proceeds "'below" the surfactant. This type of epitaxial growth is techni- cally very important, because it may reduce the transition temperature between 2D growth and step flow growth modes and may suppress the 3D island formation which occurs on the surface without the surfactant. It is believed that the surfaetant reduces sur- face energies of thc substrate and the overgrown material (which may be the same material in the case of homoepitaxy as in the present study). Another important point is that during growth the substratc temperature must be high enough so that bond breaking between the surfaetant and the substrate (or the overgrowth) and rebonding between the surfactant and the overgrowth can take place. The atomic processes of this type of growth, however, have not been made clear. We have observed growth of Si and Ge on a Si(I 1 l)5 × 2-Au surface by reflection electron mi- croscopy (REM). The 5 × 2 structure transforms to the disordered 1 x 1 structure at about 750°C [4,5] and adsorbate Au atoms disappear from the surface above 800°C. This indicates that the en- ergy for the bond breaking between a Si and a Au atom is rather large. Thus, Au may not be a good candidate of the surfactant on Si. However, there are several reasons for the choice of Au: (l) a small amount of Au deposit does not diffuse into the bulk even at high temperatures; (2) a Au deposit transforms the complicated 7 x 7 struc- ture to the rather simple 5 × 2 structure with a low Au coverage of about 0.5 monolaycr (ML) [6]; (3) the 5 × 2 structure introduces new surface defects such as domain boundaries, which may act as preferential nucleation sites of 2D islands; and (4) orientational relations between the 5 × 2 domains on the 2D islands and the 5 × 2 domains of the substrate in the case of Si deposition are interesting problems and such restructuring proc- esses are important in surfactant-mediated growth. The present REM study shows some details of these growth processes. Here, the case of Si growth is reported. The growth of Ge will be reported elsewhere [7]. 2. Experimental An ultra-high-vacuum electron microscope equipped with an evaporator for metals and an 11169_4332/I12/5ll5,t}11 r!~ lgLi2 - Elsevier Science Publishers B.V. All rights reserved