A Computational Platform to Unravel the Mechanisms of Directed Self-Assembly of Nanoparticles on Surfaces

The development of miniaturization in semiconductor manufacturing is mainly driven by the continuous pursuit of extremely powerful performance and efficient energy consumption behind which requires more sophisticated design of devices in nanoscale, even beyond sub-10 nm dimension. Integration of nanomaterials with such patterned devices would further promote cooperative phenomenon with adjacent substrate materials, and demonstrate extraordinary physicochemical properties. However, the fabrication process of above-mentioned low-dimensional devices demands strict control on nanostructure positioning that incurs several difficulties. None of the conventional patterning techniques with top-down photolithography is able to adequately address the challenge due to physical restriction in resolution. Alternative techniques have been introduced to solve the problem, including ultra-violet lithography, nanoimprint lithography, and directed self-assembly, among which directed self-assembly presents a potential in producing structurally and functionally complex substrate-supported nanostructure in a well-tunable and cost-effective manner, in addition to its broad applicability and operational simplicity. Recent development in directed self-assembly on templated surfaces, while remains at a beginning stage, has presented the opportunity of assembling arrays of nanoparticles. Advanced control over the assembly process including the dynamics of advancing/receding contact line, optimization of con finement and template, nevertheless, requires accurate and complete understanding of the underlying mechanisms. In addition, the experimental study on the contact line behavior or further in directed self-assembly needs to create a thin lm of hundred nanometers thickness which remains a challenge in terms of efficiency and cost. Nevertheless, computational modeling provides with great feasibility and convenience to investigate a variety of parameters at specific conditions. We develop a 3D kinetic Monte Carlo model to understand the dynamics of the liquid film developed during dip coating process and the wetting behavior of a liquid lm rising along a vertical substrate. By varying gravitational acceleration and surface tension, we present explicit analysis of the effect of dynamics of solid-liquid interface on the interfacial displacement and contact line roughness. To further unravel the wetting mechanism of spontaneous rise of thin liquid lm, we present a coarse grained Molecular Dynamics simulation. We investigate the dynamics of a rising contact line by demonstrating its displacement and dynamic contact angle in single- and double-wall geometry, and different surface roughness. Lithographically-de ned substrates also play a signifi cant role in pinning, deformation, and contact line of a liquid lm receding over the substrate. Using a coarse grained Molecular Dynamics simulation, we show the pinning force and distortion of the pinned contact line varies across different nanocavity shapes and orientation. A rotational flow at the receding contact line is observed, and a localized clamping effect originated from the variation of dynamic contact angle is determined, and discussed. To explore the impact of confinement, nanoparticle density, and template geometry on the directed self-assembly of nanoparticles, we utilize a series of coarse grained Molecular Dynamics simulations. We develop a phase diagram for the directed self-assembly of nanoparticles. From the phase diagram, we show that high yield of nanoparticle deposition is obtained at specific combination of liquid film thickness (con finement) and nanoparticle density. We propose a new mechanism for the directed self-assembly by providing a series of analyses of nanoparticles trajectories. The new mechanism has roots in random hopping between the bulk liquid and nanocavity, which is shown to occur far from liquid-vapor meniscus. We also discuss the impact of template geometry on the yield by modifying the center-to-center distance and circular nanocavity radius. We then turn to the energetics of the directed self-assembly, and determine the free energy and entropic contributions to the self-assembly of nanoparticles, using Jarzynski's equality. Taking advantage of the Steered Molecular Dynamics simulations with both stagnant and receding liquid films, we discuss the impact of nanoparticle densities and receding contact line on the free energy contributions. We show that the directed self-assembly of nanoparticles is entropically prohibitive at low nanoparticle densities, and energetically unfavorable at high nanoparticle densities. Last but not least, we propose future directions, which can facilitate the template design, controlling non-equilibrium process of directed self-assembly of nanoparticles, and optimizing the yield during experiments, using machine learning with a continuous feedback loop linked to the experiments.