Search for: All records

Thanks to their softness, biocompatibility, porosity, and ready availability, hydrogels are commonly used in microfluidic assays and organ-on-chip devices as a matrix for cells. They not only provide a supporting scaffold for the differentiating cells and the developing organoids, but also serve as the medium for transmitting oxygen, nutrients, various chemical factors, and mechanical stimuli to the cells. From a bioengineering viewpoint, the transmission of forces from fluid perfusion to the cells through the hydrogel is critical to the proper function and development of the cell colony. In this paper, we develop a poroelastic model to represent the fluid flow through a hydrogel containing a biological cell modeled as a hyperelastic inclusion. In geometries representing shear and normal flows that occur frequently in microfluidic experiments, we use finite-element simulations to examine how the perfusion engenders interstitial flow in the gel and displaces and deforms the embedded cell. The results show that pressure is the most important stress component in moving and deforming the cell, and the model predicts the velocity in the gel and stress transmitted to the cell that is comparable to in vitro and in vivo data. This work provides a computational tool to design the geometry and flow conditions to achieve optimal flow and stress fields inside the hydrogels and around the cell. 

Abstract Polymer materials suffer mechano-oxidative deterioration or degradation in the presence of molecular oxygen and mechanical forces. In contrast, aerobic biological activities combined with mechanical stimulus promote tissue regeneration and repair in various organs. A synthetic approach in which molecular oxygen and mechanical energy synergistically initiate polymerization will afford similar robustness in polymeric materials. Herein, aerobic mechanochemical reversible-deactivation radical polymerization was developed by the design of an organic mechano-labile initiator which converts oxygen into activators in response to ball milling, enabling the reaction to proceed in the air with low-energy input, operative simplicity, and the avoidance of potentially harmful organic solvents. In addition, this approach not only complements the existing methods to access well-defined polymers but also has been successfully employed for the controlled polymerization of (meth)acrylates, styrenic monomers and solid acrylamides as well as the synthesis of polymer/perovskite hybrids without solvent at room temperature which are inaccessible by other means. 

Hexagonal semiconductors such as 4H SiC have important high-frequency, high-power, and high-temperature applications. The applications require accurate knowledge of both ordinary and extraordinary relative permittivities, ε and ε||, perpendicular and parallel, respectively, to the c axis of these semiconductors. However, due to challenges for suitable test setups and precision high-frequency measurements, little reliable data exists for these semiconductors especially at millimeter-wave frequencies. Recently, we reported ε|| of 4H SiC from 110 to 170 GHz. This paper expands on the previous report to include both ε and ε|| of the same material from 55 to 330 GHz, as well as their temperature and humidity dependence enabled by improving the measurement precision to two decimal points. For example, at room temperature, real ε and ε|| are constant at 9.77 ± 0.01 and 10.20 ± 0.05, respectively. By contrast, the ordinary loss tangent increases linearly with the frequency f in the form of (4.9 ± 0.1)  10−16 f. The loss tangent, less than 1  10−4 over most millimeter-wave frequencies, is significantly lower than that of sapphire, our previous low-loss standard. Finally, both ε and ε|| have weak temperature coefficients on the order of 10−4 /°C. The knowledge reported here is especially critical to millimeter-wave applications of 4H SiC, not only for solid-state devices and circuits, but also as windows for high-power vacuum electronics. 

Counterfactual explanations of Graph Neural Networks (GNNs) offer a powerful way to understand data that can naturally be represented by a graph structure. Furthermore, in many domains, it is highly desirable to derive data-driven global explanations or rules that can better explain the high-level properties of the models and data in question. However, evaluating global counterfactual explanations is hard in real-world datasets due to a lack of human-annotated ground truth, which limits their use in areas like molecular sciences. Additionally, the increasing scale of these datasets provides a challenge for random search-based methods. In this paper, we develop a novel global explanation model RLHEX for molecular property prediction. It aligns the counterfactual explanations with humandefined principles, making the explanations more interpretable and easy for experts to evaluate. RLHEX includes a VAE-based graph generator to generate global explanations and an adapter to adjust the latent representation space to human-defined principles. Optimized by Proximal Policy Optimization (PPO), the global explanations produced by RLHEX cover 4.12% more input graphs and reduce the distance between the counterfactual explanation set and the input set by 0.47% on average across three molecular datasets. RLHEX provides a flexible framework to incorporate different human-designed principles into the counterfactual explanation generation process, aligning these explanations with domain expertise. The code and data are released at https://github.com/dqwang122/RLHEX.