The morning session of FNANO consisted of a suite of presenters who discussed their photonics and self-assembly techniques. Ralf Jungmann talked about the DNA-PAINT technique which I find really remarkable. The idea is strange: transient interactions + fuzzy pictures + math = sharp pictures.
One can image a single molecule with fluorescence microscopy (this has been done for 20 years) but the details get lost. It just looks like a cloud and it doesn’t matter how good a microscope you use. It will always be a cloud. DNA-PAINT is designed to generate such clouds over and over. By carefully analyzing the series of transient fuzzy clouds, a computer can find their exact centers. By comparing the centers of all of the clouds (which overlap but appear at different times; I said it was weird) the computer can construct an image that is waaay below the limit of normal microscopes. I love it and want to try it.
Yan Liu and Philip Tinnefeld both talked about using carefully arranged metal structures to have a strong impact on how light interacts with matter. For instance, two metal nanospheres on either side of a fluorescent molecule make that molecule considerably brighter – like hooking it up to an antenna. Some thought was given to the relationship between humans new entry into engineering on the scale of light-waves, and how biology has been doing it for billions of years to accomplish photosynthesis. I think it will be hard to compete with nature in this arena, but I think it’s a wonderful enterprise.
In the afternoon session I got to hear mostly about non-biological self-assembly. One talk by Lee Cronin stood out. He makes nano-scale objects that are not made from DNA. That was an interesting change. Almost everything else has been DNA. His structures are not as designable in terms of shape and tend to be much more symmetrical. Still, his approach holds special appeal to me. He made a liquid handling robot out of a 3D printer that iterated hundreds of experimental self-assembly conditions in order to find an optimum. I did something similar a few years ago. I think that robotic approaches to the experimental work are a great way to do science without going crazy. Robots can generate a rich, reproducible dataset that can then be mined for interesting features.