I think it is safe to say that directed self-assembly was the most striking advance reported at the recent Advanced Lithography Symposium. This self-assembly is called directed because it needs a bit of help in the form, say, of a pre-patterned surface in order to take on defect-free long range order. For the hard disk market, directed self-assembly may well enable bit densities above 1 Tb / sq. in. [See US20090196488A1 and Park et al.]
One may also forsee a combination of interference-assisted lithography and directed self-assembly leading to nanometer-scale active layer patterning of semiconductors. But does self-assembly presage a more fundamental shift in nano-structuring, leading to 2D arrangements that IC lithography has overlooked? And to 3D arrangements of materials? Almost certainly.
Some time ago, Zheng et al., writing in PNAS, demonstrated the self-assembly of LED segments onto device carriers in a turbulent liquid medium. In another example, Stauth and Parviz demonstrated self-assembly of active devices (FETs) onto plastic substrates, thereby forming logic circuits.
The examples are admittedly at the micro-scale rather than the nano-scale, but significant scale-independent work is on-going. Torquato and colleagues have produced a remarkable series of results by designing isotropic interaction potentials so that particles self-assemble into square, cubic, honeycomb, and diamond lattices. Take your pick. Torquato presents a general methodology in this article.
Another approach to designer interaction potentials by du Toit et al. yields more complex structures including the kagome lattice. Even more exciting is the very recent exploration of the kinetics of self-assembly by Pankavich et al. Their approach should readily accomodate any designer potential, permitting not only the prediction of the equilibrium (ground) state, but also telling us something about its stability.
As we know, the world is not flat. Bowick and Giomi published in 2009 an extensive review of two-dimensional systems on curved surfaces. (Consider the surfaces of nanotubes or of buckyballs, for example, or softer surfaces like those in liquid/liquid colloids.) One might think that this is rather far removed from the flat surfaces of the nanolithographic world. Apart from certain topological constraints (the bald spots on a hair-covered sphere, for example), it turns out that the relationship to flat surfaces is remarkably close. Indeed, Cohn and Kumar not only demonstrate a deep connection but also establish an algorithm leading to greatly simplified interaction potentials. This latter article has strong mathematical foundations; I anticipate that the method will yield robust results. That said, the paper is brief and eminently readable.