(359e) Prediction of Self-Assembly Propensity for Patchy Particles: Pathways and Fingerprints

Choosing the shapes and interactions between Brownian building blocks such that they self-assemble a desired structure is difficult because of the large available parameter space. In this work we predict the self-assembly propensity for a set of patchy particles by using bottom up building block assembly to generate their free-energy minimizing assembly pathways. Using a shape-matching algorithm we identify kinetic traps inconsistent with a desired target pattern and generate a convenient visualization that predicts assembly propensity at a particular temperature. Quickly generating these assembly pathway "fingerprints" we identify the building blocks and experimental conditions that optimize the self-assembly of desired structures without more expensive experiments or simulations. The predictive capabilities of the pathway fingerprints are confirmed with cluster Monte Carlo simulations and we discuss the problem domains where such methods are best suited.