Data-Driven Prediction of Janus/Core-Shell Morphology in Polymer Particles: A Machine-Learning Approach

Engineering multifunctional surface topography to regulate multiple biological responses

Surface topography or curvature plays a crucial role in regulating cell behavior, influencing processes such as adhesion, proliferation, and gene expression. Recent advancements in nano- and micro-fabrication techniques have enabled the development of biomimetic systems that mimic native extracellular matrix (ECM) structures, providing new insights into cell-adhesion mechanisms, mechanotransduction, and cell-environment interactions. This review examines the diverse applications of engineered topographies across multiple domains, including antibacterial surfaces, immunomodulatory devices, tissue engineering scaffolds, and cancer therapies. It highlights how nanoscale features like nanopillars and nanospikes exhibit bactericidal properties, while many microscale patterns can direct stem cell differentiation and modulate immune cell responses. Furthermore, we discuss the interdisciplinary use of topography for combined applications, such as the simultaneous regulation of immune and tissue cells in 2D and 3D environments. Despite significant advances, key knowledge gaps remain, particularly regarding the effects of topographical cues on multicellular interactions and dynamic 3D contexts. This review summarizes current fabrication methods, explores specific and interdisciplinary applications, and proposes future research directions to enhance the design and utility of topographically patterned biomaterials in clinical and experimental settings.

Glass Transition Temperature Prediction of Polymers via Graph Reinforcement Learning

An expansive array of graph-based models has been utilized for accurate prediction of the structure-property relation of polymers. However, these approaches notably underutilize unsupervised structural information. Concentrating on the domain of heterocyclic polymers, particularly polyimides, this study delves into the glass transition temperature (Tg) prediction, aiming to fully exploit the potential within both the global and local structures of molecules. To achieve this, a graph reinforcement learning framework termed Molecular Structural Regularized Graph Convolutional Network with Reinforcement Learning (MSRGCN-RL) is proposed. Experimental results highlight the crucial role of both global and local structural regularization in precise Tg prediction. Concurrently, optimization of MSRGCN training through RL proves essential. This research leads the way in integrating Graph Neural Networks (GNNs) with reinforcement learning methodologies for the property prediction of polymers.

A unified thermodynamic and kinetic approach for prediction of microcapsule morphologies

HYPOTHESIS

Microcapsule formation, following internal phase separation by solvent evaporation, is controlled by two main factors of thermodynamic and kinetic origin. Morphology prediction has previously focused on the final thermodynamical state in terms of spreading conditions, limiting the prediction accuracy. By additionally considering kinetic effects as the emulsion droplet evolves through the two-phase region of its ternary phase diagram during solvent evaporation, this should enhance prediction accuracy and explain a wider range of morphologies.

EXPERIMENTS

Dynamical interfacial tensions, and thereby spreading coefficients, during the formation of poly(methyl methacrylate) and poly(d,l-lactic-co-glycolic acid) microcapsules were measured by first establishing the boundaries and tie-lines of the ternary system in the emulsion droplets. Kinetic effects during the formation were investigated by varying the solvent evaporation rate and hence the time for polymer shell formation equilibration. The theory was validated by comparing predicted morphologies to microscopic snapshots of intermediate and final morphologies.

FINDINGS

The proposed theory explained both intermediate acorn and core-shell morphologies, where a late transition from acorn to core-shell produced microcapsules containing highly off-centered cores. By considering the kinetic factors, the formulation could be altered from yielding kinetically frozen acorns to core-shell and from yielding multicore to single core microcapsules.

Tailored Morphology in Polystyrene/Poly(lactic acid) Blend Particles: Solvent's Effect on Controlled Janus/Core-Shell Structures

Controlling the morphology of polymeric particles is vital for their diverse applications. In this study, we explored how solvent composition influences the morphology of poly(styrene)/poly(lactic acid) (PS/PLA) particles prepared via the emulsion solvent evaporation method. We used toluene, dichloromethane (DCM), and various mixtures to prepare these particles. We investigated phase separation within the PS/PLA/solvent system using the Flory-Huggins ternary phase diagram and MesoDyn simulation, revealing pronounced immiscibility and phase separation in both PS/PLA/DCM and PS/PLA/toluene systems. We employed scanning electron microscopy (SEM) and attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR) to characterize the resulting morphologies. Our study unveiled the substantial impact of solvent composition on particle structure. Using pure toluene resulted in acorn-shaped Janus particles. However, incorporating DCM into the solvent induced a transition from Janus to core-shell morphology. Remarkably, core-shell particles exhibited a single-core structure in a mixed toluene/DCM solvent, indicating thermodynamic stability. In contrast, pure DCM favored kinetically controlled multicore morphology, leading to lower PLA crystallinity due to increased PS-PLA interfaces. Samples with high Janus balance formed a self-assembled, two-dimensional (2-D) monolayer film, demonstrating the interfacial activity of the Janus particles. This 2-D monolayer film exhibits desirable emulsification properties with potential applications in various fields. Our study combines theoretical and experimental analyses, shedding light on the profound impact of solvent composition on the PS/PLA particle morphology. We observed transitions from Janus to core-shell structures, highlighted the influence of solvent viscosity on particle size, and uncovered the formation of self-assembled 2-D monolayer films. These insights are pivotal for tailoring polymeric particle structures. Furthermore, our findings advance macromolecular science in interface design, offering promising prospects for innovative materials development.