The initial stages of the development of a Ge adatom layer on a clean Si(00
1)-(2 x 1) surface are consistent with random deposition and limited surfac
e mobility. A critical comparison of the rate of attenuation of the Si buck
led dimer up-atom signal by Ge deposition with the growth of the two well-r
esolved features of the Ge adatom spectrum rules out the substitution of Ge
into the Si dimers. Instead, Ge is captured by the dangling bonds of the S
i dimers and remains on the surface, initially dominantly as isolated Ge at
oms, then as dimers, and finally in islands or clusters. (C) 2000 Elsevier
Science B.V. All rights reserved.