Programmable Control of Nucleation for Algorithmic Self-Assembly

• DOI: 10.1137/070680266
• Publication Date: 2009-12-04T23:38:28Z
• Authors: Rebecca Schulman; Erik Winfree
• Keywords: Spurious relationship

Algorithmic self-assembly, a generalization of crystal growth processes, has been proposed as mechanism for autonomous DNA computation and bottom-up fabrication complex nanostructures. A "program" growing desired structure consists set molecular "tiles" designed to have specific binding interactions. key challenge making algorithmic self-assembly practical is designing tile programs that make assembly robust errors occur during initiation growth. One method the controlled assembly, often seen in biology, use seed or catalyst molecule reduces an otherwise large kinetic barrier nucleation. Here we show how program similarly, such seeded proceeds quickly but there arbitrarily unseeded We demonstrate this technique by introducing family sets which rigorously prove that, under right physical conditions, linearly increasing size exponentially rate spurious Simulations these "zig-zag" suggest plausible experimental it possible grow crystals just few hours less than 1 percent are spuriously nucleated. Simulation results also zig-zag could be used detection single strands. Together with prior work showing can made properly initiated growth, demonstrates objects via proceed both efficiently low error rate, even model where local rules probabilistic.