Nonequilibrium Structures of C₆₀ on CaF₂(111): Exploring Structural Variability by Different Sample Preparation Pathways

Abstract

Molecular self-assembly is considered a promising tool for creating functional molecular structures on surfaces, e.g., for future molecular electronic devices and sensor applications. In molecular self-assembly, the target structure is encoded in the molecular building blocks, which are driven toward their thermodynamically favored arrangement on the surface. This approach is, however, intrinsically limited to a single molecular pattern on the surface. Structures beyond this thermodynamic ground state minimizing the free energy might become accessible through competing pathways. Here, we make use of different sample preparation pathways for arriving at distinctly different molecular structures of C 60 on the (111) cleavage plane of the calcium fluoride. Using dynamic atomic force microscopy operated in an ultrahigh vacuum, we investigate the resulting island geometries as a function of the preparation pathway. When deposited onto the surface at low temperatures (about 120 K), C 60 forms single-layer hexagonal islands. Upon heating to about 320 K, these islands transform into double-layer islands with irregular edges (two-step experiment). Interestingly, distinctly different, truncated triangular double-layer islands are obtained from a direct preparation pathway (one-step experiment), i.e., when the molecules are deposited onto the sample kept at 320 K. This pathway dependence demonstrates that nonequilibrium structures are involved. Our results are corroborated by kinetic Monte Carlo simulations, which reveal the same pathway-dependent structures as those in the experiments. Based on the simulations, we identify the barrier for freely diffusing molecules to jump from the first into the second layer (ascension barrier) as key for the formation of the different island morphologies. The second ingredient to arrive at different islands is that edge diffusion rates for single-layer and double-layer islands differ. While edge diffusion is enabled for single-layer islands, it is greatly suppressed in the case of double-layer islands (essentially due to additional bonds formed with molecules in the second layer). Thus, a given island shape is effectively stabilized when molecules can jump into the second layer. In essence, allowing the molecules to ascend into the second layer provides a means to stabilize the prepared molecular islands. Our work illustrates how different preparation protocols can be used to enhance the structural variability of molecular structure formation on surfaces.