Molecular flasks are compounds that are able to mediate or catalyze chemical transformations inside their cavities. The development of such compounds is often inspired by nature. Enzymes, nature's catalysts, are able to convert a certain substrate with very high turnover number and selectivity. In addition to their very high chemo-, regio-, and stereoselectivity, enzymes are also able to distinguish their substrates on the basis of size, resulting in size selectivity. To date, many synthetic materials such as metal− organic frameworks are used to accomplish size-selective transformations. However, also the number of molecular flasks known to mediate or catalyze size-selective transformations is increasing. In this perspective an overview on classic and the most recent examples of size-selective molecular flasks is given. In addition, an outlook on promising developments in cavity chemistry that may lead to the development of additional sizeselective molecular flasks is given.

A cobalt-porphyrin catalyst encapsulated in a cubic M8 L6 cage allows cyclopropanation reactions in aqueous media. The caged-catalyst shows enhanced activities in acetone/water as compared to pure acetone. Interestingly, the M8 L6 encapsulated catalyst reveals size-selectivity. Smaller substrates more easily penetrate through the pores of the "molecular ship-in-a-bottle catalysts" and are hence converted faster than bigger substrates. In addition, N-tosylhydrazone sodium salts are easy to handle reagents for cyclopropanation reactions under these conditions.

The synthesis of a new, cubic M8L6 cage is described. This new assembly was characterised by using NMR spectroscopy, DOSY, TGA, MS, and molecular modelling techniques. Interestingly, the enlarged cavity size of this new supramolecular assembly allows the selective encapsulation of tetra(4-pyridyl)metalloporphyrins (M(II)(TPyP), M = Zn, Co). The obtained encapsulated cobalt-porphyrin embedded in the cubic zinc-porphyrin assembly is the first example of a catalytically active encapsulated transition-metal complex in a cubic M8L6 cage. The substrate accessibility of this system was demonstrated through radical-trapping experiments, and its catalytic activity was demonstrated in two different radical-type transformations. The reactivity of the encapsulated Co(II)(TPyP) complex is significantly increased compared to free Co(II)(TPyP) and other cobalt-porphyrin complexes. The reactions catalysed by this system are the first examples of cobalt-porphyrin-catalysed radical-type transformations involving diazo compounds which occur inside a supramolecular cage.

Cobalt(III)–carbene radicals, generated through metalloradical activation of salicyl N-tosylhydrazones by cobalt(II) complexes of porphyrins, readily undergo radical addition to terminal alkynes to produce salicyl-vinyl radical intermediates. Subsequent hydrogen atom transfer (HAT) from the hydroxy group of the salicyl-moiety to the vinyl-radical leads to the formation of 2H-chromenes. The Co(II)-catalyzed process can tolerate various substitution patterns and produces the corresponding 2H-chromene products in good isolated yields. EPR spectroscopy and radical-trapping experiments with TEMPO are in agreement with the proposed radical mechanism. DFT calculations reveal the formation of the salicyl-vinyl radical intermediate by a metalloradical mediated process. Unexpectedly, subsequent HAT from the hydroxy moiety to the vinyl radical leads to formation of an o-quinone methide intermediate, which dissociates spontaneously from the cobalt center and easily undergoes an endo-cyclic, sigmatropic ring-closing reaction to form the final 2H-chromene product.

Encapsulation of a manganese porphyrin in a self-assembled molecular cage allows catalytic epoxidation of various substrates in 1:1 water/acetonitrile mixtures. The cage acts as a phase-transfer catalyst and creates a protective environment for the catalyst improving the stability. The encapsulated catalyst also allows discrimination between styrene derivatives of various sizes. In a direct competition experiment, the selectivity of the epoxidation reaction could be inverted with respect to a benchmark catalyst.