The Raizen group at UT Austin recently proposed [arXiv:1004.5581] a nano-lithography method involving masked, magnetically focussed, neutral atomic beams. This involves building a mask from a magnetic film, suspending it at some species- and condition-dependent focal distance from the substrate, magnetizing it, and spraying it with a supersonic atomic beam.
In their strictly theoretical paper, they calculate that one could deposit 18 nm indium dots, for example. Interestingly, the focal length in this case is about 82 microns. Working from their assumptions, one can calculate a thickness deposition rate of about 0.1 nm per second. As might be expected, lighter elements show shorter focal lengths for the same magnetization.
Practical difficulties abound. For example, among the assumptions is that the holes in the mask are 150 nm in diameter. This rather strongly limits the pattern pitch. Nevertheless, the community should embrace this kind of “out of the box” thinking. How else would we have gotten directed self-assembly?