Designing Bottlebrush Networks: Where Theory, Simulations, and AI Meet Forensics

Designing polymer networks with programmable mechanical properties is essential for advancing materials for medical implants, tissue engineering, soft robotics, and wearable electronics. In this talk, I will show how viscoelastic properties of soft materials can be encoded in synthetic elastomers by tailoring the molecular architecture of brush-like network strands. The developed framework integrates theory, computer simulations, synthesis and experiments to discover correlations between mechanical properties and architecture of brush networks. At its core is a packing parameter defined in terms of the effective Kuhn length and occupied volume of the network strands. This universal descriptor enables a precise mapping of mechanical properties onto molecular parameters: the degree of polymerization and grafting density of side chains, and the degree of polymerization of the brush backbones between crosslinks. This design strategy of soft materials is tested by reproducing mechanical properties of diverse biological gels and tissues using poly(dimethyl siloxane) and poly(n-butyl acrylate) elastomers with brush-like strands. The emerging hierarchical structure-property correlations in these systems open new opportunities for combining theory-driven design with AI-based approaches to encode targeted mechanical responses. I will conclude by outlining a “forensic” strategy for unveiling network structure from its nonlinear stress-strain signatures.