Shape Selection in Nanopillar Formation

Abstract

Crystal growth processes produce a diverse array of surface formations, primarily distinguished by their geometric shapes. While some structures strictly adhere to the underlying crystal symmetry, others exhibit universal circular or oval geometries. Utilizing Vicinal Cellular Automata (VicCA) modeling, we demonstrate that these morphological differences depend on the spatial distribution of the growth potential. Specifically, local potential variations concentrated around surface steps drive the formation of the lattice symmetry - following structures, whereas global potentials - often originating from defects-generate universal spherical or oval shapes. Furthermore, we illustrate how these morphologies are influenced by the growth parameters such as sticking coefficient or diffusion coefficient. Although the positioning of surface defects is difficult to control, we show that temperature and external particle flux can be effectively used to steer and manipulate surface pattern formation.

Figures (4)

• Figure 1: Visualization of the potential landscape. Side view of the surface with the effective a) local and b) global potential.
• Figure 2: A hexagonal nanopillar obtained through adatom diffusion governed by local potential $\beta E_V=4$, $\beta E_{ES}=2$, $n_{DS}=10$, $c_0=0.005$. Number of simulation steps $2^{.}10^6$.
• Figure 3: Nanopillars obtained by adatom diffusion, driven by a cylindrically shaped global potential, with the probability of an adatom attaching to the kink position is given by $p_k=e^{-\Delta}$, and $\Delta=$ a) -3.5 b) -2.2 and c) -1.2. Probability of attachment to the step $p_s=p_k^2$, $n_{DS}=10$, $c_0=0.01$. Number of simulation steps $2^{.}10^6$.
• Figure 4: Nanopillars obtained through adatom diffusion, governed by global potential with an ellipsoidal shape. Probability of an adatom attaching to the kink position is given by $p_k=e^{-\Delta}$, and $\Delta=-2.2$. Other parameters are the same as in Fig. \ref{['fig:results-global']}