When Glass and Gel Meet

Chaudhuri, P.¹, Hurtado, P.², Berthier, L.³ and Walter, K.³

¹ Universite Claude Bernard Lyon 1, France; ² Institute "Carlos I" for Theoretical and Computational Physics, Universidad de Granada, Spain; ³ Université Montpellier II, France

The structure and relaxation dynamics of dense glass-formers is nowadays relatively well-understood. Their behavior is characterized by a dramatic slowing down of the dynamics due to steric hindrance effects: particles are arrested within the cage formed by their neighbors. On the other hand, low-density frozen disordered structures, i.e. gels, are also observed in Nature. In this case particles become arrested when bonded to a macroscopic network structure which percolates through the system. The relaxation dynamics is less well-understood in gels than in dense glass-formers, and one finds a richer zoology. Certain systems can be studied both at low and high densities, and thus one can use them to investigate the glass and gel phases. What exactly happens at intermediate densities at which the two phases meet is not clear.

Here we numerically investigate a simple model system that allows to study the gel and the dense glass state, and hence to shed light on the interplay of two different arrest mechanisms. The model, first introduced in Phys. Rev. Lett. 98, 135503 (2007), is composed by particles and bridging polymers, and has a low-density equilibrium gel phase. The global structure of the gel is homogeneous, but the stress is only supported by a fractal network. As the density is increased, the systems enters into a glassy phase which interferes with the existing gel structure. In this regime the system exhibits a three-step relaxation process as a result of the competition between two different localization lengthscales: l_SH (associated to steric hindrance effects) and l_PN (related to the mesh size of polymer network). In some regions of parameter space this competition gives rise to an apparent logarithmic decay of correlation functions, which can be understood as a simple crossover effect. We propose a theoretical model that describes the observed behavior, sheding light on the origin of this lengthscales competition and opening the door to the investigation of a very large class of glass-forming systems with new and rich dynamics.