Search for: All records

Self-assembly of amphiphilic molecules is an important phenomenon attracting a broad range of research. In this work, we study the self-assembly of KTOF4 sphere–rod amphiphilic molecules in mixed water–dioxane solvents. The molecules are of a T-shaped geometry, comprised of a hydrophilic spherical Keggin-type cluster attached by a flexible bridge to the center of a hydrophobic rod-like oligodialkylfluorene (OF), which consists of four OF units. Transmission electron microscopy (TEM) uncovers self-assembled spherical structures of KTOF4 in dilute solutions. These spheres are filled with smectic-like layers of KTOF4 separated by layers of the solution. There are two types of layer packings: (i) concentric spheres and (ii) flat layers. The concentric spheres form when the dioxane volume fraction in the solution is 35–50 vol%. The flat layers are formed when the dioxane volume fraction is either below (20 and 30 vol%.) or above (55 and 60 vol%.) the indicated range. The layered structures show no in-plane orientational order and thus resemble thermotropic smectic A liquid crystals and their lyotropic analogs. The layered packings reveal edge and screw dislocations. Evaporation of the solvent produces a bulk birefringent liquid crystal phase with textures resembling the ones of uniaxial nematic liquid crystals. These findings demonstrate that sphere–rod molecules produce a variety of self-assembled structures that are controlled by the solvent properties. 

Prior to fracture, a polyacrylamide hydrogel has a stress-stretch curve of nearly perfect elasticity, but it has been suggested that an inelastic zone exists around a crack tip. This inelastic zone, however, has never been observed directly in a polyacrylamide hydrogel. Here we identify the inelastic zone using digital image correlation (DIC). We prepare a polyacrylamide hydrogel with a precut crack. While a sample of the hydrogel is stretched, the speckle patterns are recorded using a microscope or a camera, with pixel size 2.3 μm and 22.7 μm, respectively. The speckle patterns recorded by the microscope and camera are processed using the DIC software, and merged to provide the deformation field over the entire sample. The measured field of deformation is used to calculate the field of energy density according to the neo-Hookean model. When the body is perfectly elastic, the field of energy density around the crack tip is inversely proportional to the distance from the crack tip. The difference between the measured field and the predicted elastic field identifies the inelastic zone. The measured size of the inelastic zone is ∼ 0.6 mm. We further confirm that, when a sample is much larger than the inelastic zone, an annulus exists, in which the elastic crack tip field prevails. 

Abstract Training language-conditioned policies is typically time-consuming and resource-intensive. Additionally, the resulting controllers are tailored to the specific robot they were trained on, making it difficult to transfer them to other robots with different dynamics. To address these challenges, we propose a new approach called Hierarchical Modularity, which enables more efficient training and subsequent transfer of such policies across different types of robots. The approach incorporates Supervised Attention which bridges the gap between modular and end-to-end learning by enabling the re-use of functional building blocks. In this contribution, we build upon our previous work, showcasing the extended utilities and improved performance by expanding the hierarchy to include new tasks and introducing an automated pipeline for synthesizing a large quantity of novel objects. We demonstrate the effectiveness of this approach through extensive simulated and real-world robot manipulation experiments.