Inverse Design of Complex 3D Nanostructures of Block Copolymers

Directed self-assembly (DSA) of block copolymers (BCP) provides a powerful tool to fabricate complex 2D nanopatterns at small length scale. Fabrication of complex 3D nanostructures by BCP, however, remains a challenge. Here, we introduce a method based on self-directed self-assembly, that utilizes artificial intelligence tools for inverse design of aperiodic 3D nanostructures. The methods presented here encompass rapid algorithms for characterizing the internal structures of BCP morphologies from the atom coordinates generated by MD simulations to substrate optimizing algorithms for developing effective routes to propagate information from the substrate into the BCP film. We will show several examples in which this AI engine is applied. In particular, we will demonstrate how to connect multilayer films or achieve novel complex bilayer structures. Our work has uncovered new design rules for novel 3D nanostructures.