Kovacs-like memory effect in dense liquids of responsive colloids

Abstract Responsive colloids (RCs) exhibit a rich non-equilibrium behavior due to the interplay between translational motion and the dynamics of internal degrees of freedom (DoFs), such as the particle size. Using Brownian dynamics (BDs) simulations of the RC-Hamiltonian performed in the isothermal–isobaric ensemble, we investigate how such a coupling of translation and dynamic size polydispersity can lead to tunable memory effects under a two-step pressure protocol inspired by Kovacs’ classic experiments. We indeed observe a substantial non-monotonic temporal relaxation of the system density in the liquid state, characteristic of Kovacs memory, whose amplitude and timing depend sensitively on how fast the internal DoF relax compared to both the translational motion and the barostat time scale. For fast barostats, we demonstrate that the memory is limited by physical system scales and thus intrinsic to the colloids. By varying the ratio of translational and internal (size) timescales overs four orders of magnitude, we demonstrate that the signatures of the Kovacs memory peak (‘Kovacs hump’) can be tuned, indicating a clear link between internal dynamics, local structure, and macroscopic memory. We hypothesize that the significant memory effect in this dense colloidal liquid (in contrast to a glassy state) originates from a combination of polydispersity, local cage relaxations, and to a significant part from the overall (average) slow size relaxation dynamics. We support the latter by demonstrating that a simple phenomenological dynamical theory based on the equation-of-state (as a function of the independent variables density and size) can qualitatively describe the BD data. Our results advance our understanding of memory effects in soft matter systems with internal configurational dynamics, opening routes for designing responsive materials with built-in memory.