Abstract
The use of a laser with a Gaussian-beam profile is frequently adopted in attempts of crystallizing nonsingle-crystal thin films; however, it merely results in the formation of polycrystal thin films. In this paper, selective area crystallization of nonsingle-crystal copper(II) oxide (CuO) is described. The crystallization is induced by laser, laser-induced crystallization, with a beam profile in the shape of a chevron. The crystallization is verified by the exhibition of a transition from a nonsingle-crystal phase consisting of small (∼100 nm × 100 nm) grains of CuO to a single-crystal phase of copper(I) oxide (Cu2O). The transition is identified by electron back scattering diffraction and Raman spectroscopy, which clearly suggests that a single-crystal domain of Cu2O with a size as large as 5 μm × 1 mm develops. The transition may embrace several distinctive scenarios such as a large number of crystallites that densely form grow independently and merge, and simultaneously, solid-state growth that takes place as the merging proceeds reduce the number of grain boundaries and/or a small number of selected crystallites that sparsely form grow laterally, naturally limiting the number of grain boundaries. The volume fraction of the single-crystal domain prepared under the optimized conditions─the ratio of the volume of the single-crystal domain to that of the total volume within which energy carried by the laser is deposited─is estimated to be 32%. Provided these experimental findings, a theoretical assessment based on a cellular automaton model, with the behaviors of localized recrystallization and stochastic nucleation, is developed. The theoretical assessment can qualitatively describe the laser beam geometry-dependence of vital observable features (e.g., size and gross geometry of grains) in the laser-induced crystallization. The theoretical assessment predicts that differences in resulting crystallinity, either single-crystal or polycrystal, primarily depend on a geometrical profile with which melting of nonsingle-crystal regions takes place along the laser scan direction; concave-trailing profiles yield larger grains, which lead to a single crystal, while convex-trailing profiles result in smaller grains, which lead to a polycrystal, casting light on the fundamental question Why does a chevron-beam profile succeed in producing a single crystal while a Gaussian-beam profile fails?