Inverse Design of Active Matter

Obiettivo

Active matter (AM) represents a fascinating class of non-equilibrium systems whereby continuous energy dissipation generates unique, dynamic behavior. For instance, when active systems like self-propelling particles (SPP) interact in large numbers they can spontaneously align and form collective phases resembling swarms or flocks of birds. In this project, we seek to inverse design (ID) i.e. to target and stabilize such spontaneous, collective phase transitions via the development of a novel ID framework. We will achieve this by combining recent developments in large deviation theory (LDT) and stochastic many-body optimization which allow us to discover and promote desired phase behavior in a general, systematic way. Chiefly, LDT allows us to exploit the natural tendencies of the system via the quantification of rare events linking the disordered to collective phase. Control forces can then be introduced and optimized to make such rare events typical. For example, in SPPs this would involve tuning torque parameters to stabilize flocking. We will apply this strategy to some representative AM models: i) deformable particle systems reproducing compression-waves in epithelium cells, and ii) field theories yielding arrested phase separation and fluid turbulence akin to those reported in assemblies of biological swimmers. We will further provide a unified picture of phase transition in AM by introducing and analyzing energetically consistent models using stochastic thermodynamics. Overall, the proposal presents an ambitious, interdisciplinary route to reveal novel strategies for ID in AM with technology implications in soft-matter systems, and to explore how energy consumption inevitably constrains the emergence of collective states in AM. This project thus combines the skills in AM from the host with the numerical and analytical experience from the fellow, setting a definitive path for the fellow to rise as an established independent researcher in statistical mechanics.