Molecular-Scale Interactions at Mineralized Collagen Interfaces Prevent Network Percolation, Preserving Compliance.

Interfaces are everywhere in technology and engineering, from electronic circuit components to structural joints and biomedical implants. Understanding and controlling these interfaces is essential for advancing device efficiency, durability, and functionality. Nature has evolved intriguing strategies for joining soft and hard tissues through the enthesis, a specialized interface between tendon and bone that exhibits an unexpected compliant region critical for interface durability. While the mechanical behavior of such biological interfaces has been partially attributed to tissue structural organization, the molecular mechanisms preventing interface stiffening, especially when filled with mineral particles, remain unknown. Using full-atomistic simulations, we discover that molecular interactions between mineral particles and collagen fibrils, the fundamental building blocks of these tissues, help maintain this essential compliance. Our models reveal that hydrogen bonds between mineral particles and collagen molecules interfere with the mineral's natural tendency to form continuous, percolated networks. This interference persists even at mineral concentrations where stiff networks would typically form in ordinary composite materials. The impact of these molecular interactions is striking: as mineral content increases, the tissue remains more compliant than predicted by traditional homogenization bounds. While individual mineral clusters create local regions of high stiffness, they are prevented from linking into a continuous load-bearing network by preferential bonding with surrounding collagen molecules. This molecular-scale control over tissue mechanics provides insight into surgical repairs and bioinspired materials for joining materials with different mechanical properties and reveals a previously unreported mechanism in polymer-matrix composites whereby molecular interactions can tune material properties by controlling network formation.