An attosecond is 10-18s. The chemistry that takes place on this timescale is called electron dynamics. For example, it is the time taken for an electron to traverse the 1s orbit in a hydrogen atom. And chemists are starting to manipulate electrons (and hence chemistry) on this timescale; for example a recent article (DOI: 10.1021/ja206193t) describes how to control the electrons in benzene using attosecond laser pulses.
The diagram above is famously attributed to Kekulé, and we now teach that this diagram represents resonance structures. This term implies electron (rather than nuclear) dynamics, which results in us only being able to observe an averaged structure with six equivalent C-C bonds. Kekulé himself (see this review) could have no way of knowing what the timescale was that his representation implied, but he certainly must have thought it was fast. What is rarely mentioned in textbooks however is how fast it actually is. If the timescale were to be less then say a nanosecond (10-9s) then we would classify such a process not as a resonance but as a valence isomerism. This does implicate nuclear motions, and such would be the appropriate description for the isomerism of say cyclo-octatetraene, which is famously slow enough to observe by NMR (see 10.1039/P29920001951). If the timescale were to be a femtosecond (10-15s) this would correspond to molecular vibrations, and benzene indeed has an observable normal vibrational mode that corresponds more or less to the representation above (the hydrogens also wag a bit). This mode even has a name, the Kekulé mode. But even this is not fast enough for the intent of the diagram above, which describes what the electrons (and not any moving nuclei) are up to.
The article I mentioned at the start, by Inga Ulusoy and Mathias Nest probes exactly this aspect at the attosecond timescale. More particularly, they look into how to shape an ultra-short laser pulse to excite the electrons into excited states of benzene. This act destroys the aromaticity of the molecule, and changes the electron dynamics in the process. I should quote them here: “We have shown that by controlling the electron dynamics we can selectively switch benzene into nonaromatic target states. These target states exhibit an ultra-fast bidirectional electron circulation around the ring system.”
This is the chemistry in (a few) attosecond(s) that I titled this blog. Whilst the article noted here is theoretical, there seems little doubt that experimental studies of chemistry in an attosecond will became more common, and who knows what surprises await us. Exciting times (sorry about the pun).
I would conclude by mentioning the other extreme, chemistry in an exasecond (1018s). This happens to correspond more or less to the age of the universe! As it happens, it is not that difficult to come up with chemical processes that occur on this timescale. Any (unimolecular) process that has a free energy barrier larger than that inferred using e.g. the equation Ln k/T = 23.76 – ΔG/RT would fit. An example might be the half life for the enantiomerisation of alanine (left to its own devices, and not interfered with by e.g. catalysts).