Stress-Induced Lamellar Order in Spider Silk Fibers

Abstract

Spider dragline silk, solely made from protein, outperforms synthetic fibers in terms of toughness and extensibility, which is thought to originate from its refined nano-scale hierarchical structure. The key for understanding the outstanding mechanical performance of dragline silk lies in a detailed description of the underlying molecular structure and dynamics under load conditions. Here, we aim at determining the ingredients for silk's outstanding toughness using multi-scale computational simulations. First, we present large-scale molecular-dynamics simulations of the structure and dynamics of silk protein. We introduce the most comprehensive, to date, spider silk models comprising of hundred 500-residue chains of two spider proteins, namely MaSp1 and MaSp2, using a collapsing-annealing protocol [1]. Our systems are composed of crystalline beta-sheet segments completely embedded in and connected with an amorphous phase. On this basis, we developed a three-dimensional continuum model of dragline silk for finite element analysis. Such model takes into account the plasticity of the beta-sheets, the rate-dependent dynamics of the amorphous phase, and the viscous friction between them [2]. Tensile properties, velocity-dependent effects and hysteresis are in good agreement with experimental data. Intriguingly, the simulations revealed that under load conditions, crystals rearrange into lamellar bands. We could confirm this trend by small-angle neutron scattering of silk fibers under stress, with simulations and experiments showing quantitatively the same shift and intensity in the signal from the periodically forming bands. The increased order results in a more homogenous stress distribution and higher fiber toughness. Our combined atomistic and fiber-level simulations and experiments suggest that the test stress-induced order is a common feature in materials combining crystalline and amorphous phases on the nano-scale.[1] Cruz-Chu, E. R et al. Faraday Discussions (2009) 143, 47-62[2] Sandeep Patil et al. PLOS One (2014) 9, e104832