Simulating the self-assembly of complex structures

Materials that can spontaneously self-assemble have been the subject of extensive recent research. It is possible to achieve a considerable degree of complexity using simple building blocks. For example, using computer simulations, we have found that 2D particles with five regularly arranged 'patches' spontaneously form dodecagonal quasicrystals in certain conditions. But whilst quasicrystals form spontaneously on cooling, it is not necessarily clear that they are also the thermodynamically stable phase. I will present a method to calculate
the free energy of the quasicrystal and use it to show that in our model system, the quasicrystal has the lowest free energy over a range of conditions and is stabilised by its greater configurational entropy over
the crystalline phases.
However, simple building blocks can only go so far and self-assembling truly 'complex' structures requires us to introduce more distinct building blocks into the system, which makes the problem of self-poisoning ever more difficult to counter. Recently, Ke and co-workers reported that DNA bricks successfully self-assembled into
structures containing not just a handful, but hundreds of distinct components [Science 338, 1117 (2012)]. However, it is not immediately obvious why such self-assembly should succeed where colloidal systems
have failed. In my talk, I will present our computational and theoretical work explaining how nucleation governs the self-assembly of these multi-component systems and the role this plays in the rational
design of the target structure.