Copper-complexed catenanes and rotaxanes in motion: 15 years of molecular machines.

In this review, we try to summarise the work performed in our laboratories in the course of the last 15 years, in the field of catenane- and rotaxane-based molecular machines containing copper(I), copper(II) or zinc(II) atoms. We put our work into perspective, showing how the properties of the compounds made have been gradually improved, mostly in terms of motion rate. In parallel, the function of the molecular machines elaborated have been made more and more complex. Instead of discussing all the systems elaborated and studied in our team, we preferred to select a few representative examples and show what were the principles which guided us for improving their performances and how the compounds were experimentally modified to afford new functions and faster-responding molecular machines. Starting from an electrochemically-driven "swinging" [2]catenane, reported in 1994, whose rearrangement was disappointingly slow, we could recently elaborate fast moving pirouetting copper-complexed [2]rotaxanes or molecular shuttles. The rearrangement mechanism of the pirouetting systems, as studied by experimental and computational methods, led us to synthesise interlocking compounds with a very open structure around the copper centre, allowing facile ligand exchange. With such compounds, whose copper centres are highly accessible, the minutes or hours required for the first generation of molecular machines to rearrange were converted to milliseconds or seconds, demonstrating that the rate limiting step in the electrochemically-steered copper catenanes and rotaxanes is probably decoordination of ligands to be replaced by entering ligands leading to the new form of the species. In addition to the electrochemically-piloted systems, we discuss a few compounds, which were not set in motion using an electrochemical signal but rather a chemical stimulus, including porphyrin-containing [2]rotaxanes or a "molecular muscle", based on a [2]rotaxane dimer of the hermaphroditic type. In these two machines, the stimulus is based on metal complexation, decomplexation or exchange (copper(I) being replaced by zinc(II) and vice versa for the "muscle"). Using non-sterically hindering ligands of the 8,8'-diaryl-3,3'-biisoquinoline, it is likely that the chemically driven motions used for contracting or stretching the "muscle" could be replaced by electrochemical signals, which are certainly more promising in terms of future devices.