Tetra-poly(ethylene glycol) (tetra-PEG) is a four-armed star polymer that can crosslink to form gels with demonstrated usefulness in various areas, ranging from separations to biomedical applications. Computational modeling is essential to understanding the behavior of tetra-PEG gels at the molecular level, enabling prediction and optimization of structures with desirable properties. However, most previous computational studies of tetra-PEG have used bead–spring coarse-grained models that are not parametrized to a specific molecular model and thus may have some challenges in making quantitative predictions that directly translate to physical experiments. Here, we systematically develop several coarse-grained tetra-PEG models from atomistic simulation data using relative entropy minimization. We design both intra- and intermolecular interactions at varying levels of coarseness using new capabilities we have implemented in our open-source software relentless. We then simulate crosslinking using our different coarse-grained models and characterize how the network properties depend on the coarse-grained model adopted. This work provides guidance about how to develop and select models for tetra-PEG (and related network formers) beyond generic bead–spring approaches.
