We have improved and evaluated hydrophilic and hydrophobic chemical treatments to prepare Si(001) and Si(111) wafers for epitaxial growth substrates. In the hydrophilic case, we found that a simple chemical treatment in a mixture of H2SO4 and H2O2 solutions has the same effect as the traditional RCA-based methods. However, the boron contamination of 5 x 10(18) atoms cm(-3) on the resultant surface after removal of its oxide layer is difficult to prove in nature and blocks surface diffusion of Si adatom at the initial stage of crystal growth. In the hydrophobic case, improvement of preparation facilities in order to free a sample from oxygen allows a surface to be obtained that is clean enough for crystal growth without the boron problem. In both cases, there is no difference in the carbon contamination concentration of 2-4 x 10(18) atoms cm(-3). These results demonstrate that the hydrophobic surface has advantages. We further studied the relationship of surface contamination and surface structures of hydrophobic Si(001) and Si(111). With increasing annealing temperature in UHV, the Si(001) surface exhibited c(4 x 4) in the narrow range of 580-750 degrees C, and eventually, a very weak SiC transmission pattern appeared above 750 degrees C. The simultaneous disappearance of c(4 x 4) and appearance of a SiC pattern indicates the association of the c(4 x 4) reconstruction with carbon. The carbon concentration of 2.5-4 x 10(18) atoms cm-3 is insufficient for carbon atoms to form a component of the c(4 x 4) structure, and therefore the carbon must have another effect, A consequence of this conclusion is that carbon is not responsible for C-type defects, and we speculate that bet-on may be the origin. The Si(111) surface showed (root 3 x root 3) at between 400 degrees C and 600 degrees C, and the reconstruction is independent of the carbon contamination. (C) 1998 Elsevier Science B.V. All rights reserved.