The main goal of self-assembly in colloidal systems is to build pre-determined structure by varying external parameters such as the effective temperature. For this to happen, the target structure has to be "encoded" in the interactions between the building blocks. Although finding interactions that favour the formation of a particular structure may seem easy just by looking at the geometry of that structure, the possibility of finding competing structures during the assembly pathways is a very difficult task. In this talk, we show how a semi-automatic procedure, based on satisfiability problem solving (a well-known class of problems in computer science) can help in predicting and avoiding competing structures. We show, both numerically and experimentally, that previously unachievable structures can be achieved through this procedure.
Emanuele Locatelli
