Liquid phase epitaxy of magnetic garnets

Abstract

In designing rare earth iron garnet compositions for magnetic bubble applications it is usually necessary to deal with no fewer than three cations and often as many as six. LPE has become the dominant process for making garnet films from PbO-B2O3-Fe2O3 fluxed melts. Flux constituents further increase the complexity of the design problem by introducing two more components. Fortunately, thermodynamics reduces the film growth problem to the simpler dimensions of a solid garnet crystallizing in a liquid fluxed melt. For epitaxy, the Pb impurity in films has a significant effect on the lattice match (fit) between a film and its substrate. A mismatch can influence film properties and the growth process as well. This factor is only of concern to the bulk crystal grower when seed crystals are employed. Garnet films can be grown on a variety of gallium garnets which have been pulled and prepared as polished wafers with high perfection only equalled in Si wafer preparation. Isothermal LPE growth is now uniformly practiced and is mainly responsible for the good reproducibility and uniformity of film products. Indeed, some of the most precise and reproducible crystal growth data ever assembled have originated from magnetic garnet studies. Interestingly, more practitioners seem to agree on what happens than on why it happens. Two kinetics models are compared by numerical methods. The relative merits of horizontal versus vertical plane dipping are discussed briefly in terms of both theory and engineering potentials. In summary, the presentation is concerned with the differences among us, the problems we face, and the apparent alternatives. © 1975.