He started his presentation by showing his first molecular square, originally prepared nearly 20 years ago [1]. This tetrameric molecule, self-assembled from Pd(II) ‘angles’ and bipyridine linkers, was the starting point of an amazing field of research. In 1995, the 2-dimensional square became a 3-d cage [2]. Again the synthesis was astonishingly simple, allowing to obtain in a high yield a nanocontainer which was found to be able to accomodate simple guests (adamantane derivatives) in its cavity. It important to realize that a ‘classical’ synthesis approach requires many more steps and complications than the simple self-assembly process devised by Fujita.

Octahedral nanocage. The vertices are palladium atoms, and four faces are occupied by tris-pyridyl ligands. Source: http://www.nature.com/materials/news/newsandviews/030522/423394a_f1.html

Interesting features of this cage are that it is water soluble (the Pd centres are charged) and its cavity can accomodate hydrophobic molecules that are not water-soluble. By varying the organic ligand, several differently shaped and sized structures were obtained by self-assembly (such as cages, bowls, boxes, tubes, catenanes, and spheres) [3].

Well it is already good to have nice-looking architectures, but then? What do you answer when your non-scientific friends asks you “So what? What is this useful for?” … Looks like people in Fujita’s group decided to tackle this question very seriously (so that they don’t have to answer ‘er well, at the moment, we don’t have practical applications but potentially it could perhaps one day blah-blah-blah…’) and thus started to investigate what their hollow macromolecules were able to do.

A striking example of application was published in 2006. The cage of the figure above was used to perform a Diels-Alder reaction between two highly hydrophobic reagents in water [4]. The first step is the inclusion of the reagents inside the cage. They are then appropriately oriented to undergo a Diels-Alder reaction, affording the final product, as indicated in the figure below. Remarkably, only one isomer is isolated (the reaction is regioselective), and it is not the isomer one obtains when the reaction is performed without the cage. Furthermore, if a bowl-shaped macromolecule is used intead of the cage to perform the same reaction, it was observed that 1) a different isomer was produced, 2) a catalytic amount of the ‘bowl’ is enough to ensure a high yield, and 3) the Diels-Alder adduct is released from the bowl after formation. This is a typically enzyme-like behaviour: once formed, the reaction product is no longer compatible with the host, and is therefore released, preventing catalyst inhibition.

Diels-Alder reactions performed inside the cage and bowl-shaped molecules.

Other impressive achievements shown during this talk included crystal engineering (a porous coordination network which is able to adapt its size and shape upon guest absorption or removal [4]) and encapsulation of biological molecules (small nucleotides duplexes formed in the hydrophobic pocket of a water-soluble cage [5]).

References:
[1] J. Am. Chem. Soc. 1990, 112, 5645. DOI: 10.1021/ja00170a042
[2] Nature 1995, 378, 469. DOI:10.1038/378469a0
[3] Acc. Chem. Rev. 2005 38 369. DOI: 10.1021/ar040153h
[4] Science 2006, 312, 251. DOI:10.1126/science.1124985
[5] Nature Chemistry 2009, 1, 53 doi:10.1038/nchem.100