ORIENTED FILM GROWTH, NOT EPITAXY IN HTSC FILM GROWTH

A survey of the literature reveals that the following materials have been used as the substrates for ''epitaxial'' deposition of YBCO high T(c) superconducting (HTSC) thin films: SrTiO3, ZrO2 (or YSZ), MgO, LaAlO3, Si and Al2O3 (with or without buffer layers), SiO2, Pt, MgAl2O4, even noncrystalline substrates. An analysis of the structural relations between the substrate and HTSC phase is presented. It is obvious that the term ''epitaxy'' is misapplied in many of these cases. We analyze the evidence for actual crystal structural control from the substrate influencing the oriented growth of the desired phase and find it nonexistent in many cases, very weak in others, and persuasive in a few cases. For the superconductor quality what matters is the quality of the HTSC film, not what it is sitting on. This quality can be quantified in four steps, the first three describing the degree of orientation in none (random-orientation), one (growth direction), or two (in film plane) crystallographic directions. The fourth is the degree of continuity of the film or absence of grain boundaries (i.e., a 100% perfectly oriented set of separated islands with grain boundaries will clearly be considerably poorer than a continuous (single) phase). We propose that instead of the term epitaxial one should focus only on the degree of orientation and continuity of the ''oriented film.'' The phase immediately contiguous to the film (including many so-called buffer layers) should be designated substrate. Any phase below such a substrate is structurally and chemically insignificant (to the YBCO growth), although it may be necessary to obtain the oriented ''buffer layer.'' Such phases should be designated as carriers, not substrates.