## ABSTRACT

Porous silicon (PSi) remains crystalline even at high porosity (p ~90%) (Unagami and Seki 1978; Duttagupta and Fauchet 1997), allowing an epitaxial growth of films of various materials. Mechanical properties of PSi strongly depend on the porosity and can be controlled by the anodization regimes (Duttagupta and Fauchet 1997) to provide PSi buffer layers with specified characteristics for the epitaxial growth. The possibility of the Si homoepitaxy on PSi to form an intercomponent isolation in ICs was patented as early as 1972 by the Nippon Telephone and Telegraph Public Corporation (Watanabe and Sakai 1972). This application dominated over a long period and was studied intensively for the fabrication of the silicon-on-insulator (SOI) structures (Bondarenko et al. 2005). Based on the Si epitaxy on PSi, Canon Inc. has developed the ELTRAN technology for the commercial production of high-quality 300-mm SOI wafers (Yonehara 2002). From the mid 1980s, PSi applicability for the heteroepitaxial growth of various films at the silicon substrate has been studied. These studies were initiated by Luryi and Suhir (1986). They theoretically predicted a possible approach for the growth of dislocation-free latticemismatched heteroepitaxial layers on a patterned substrate, such as PSi. After that, many efforts to validate the Luryi/Suhir theoretical model and to use PSi for the heteroepitaxy have been performed (Xie and Bean 1990a; Bondarenko et al. 1994, 1996; Zubia and Hersee 1999; Levchenko et al. 1999; Novikov et al. 2003; Christiansen et al. 2006; Huangfu et al. 2013; Ye and Yu 2014). Recently, research of Ge and SiGe epitaxy (Kim et al. 2007; Blanchard et al. 2011; Aouassa et al. 2012; Gouder et al. 2014a,b) also demonstrated that PSi is indeed a strong contender for specific “pseudo-substrate” applications (Dariani 2014).