At first, AIE was seen as a curiosity and deemed likely to be rare. In the 1850s, George Stokes noted that some inorganic complexes were fluorescent in the condensed, solid state but not in solution. Yet this effect of 'aggregation-induced emission' (AIE), as Tang and colleagues called it, was apparently seen, but not understood, much earlier. Aggregation-induced 'switching off' is intuitively understandable, but 'switching on' due to aggregation was more surprising. This looked like the opposite of the well-known phenomenon of concentration quenching, in which energy transfer between fluorescent (generally organic) molecules quenches the emission, an effect explained in 1955. One of the most striking instances became apparent in 2001, when Ben Zhong Tang of the Hong Kong University of Science and Technology and his co-workers found that heterocyclic silicon-containing molecules called siloles become luminescent as nanoscopic aggregates even though the individual molecules in dilute solution do not emit light.

A classic example is the appearance of bulk-like metallic behaviour in small clusters of metal atoms only once they exceed a certain critical size. Yet molecular systems do not have to be particularly complex before their properties become more than the sum of the parts.

Biology is replete with such examples, from the flocking of birds to the way metabolic processes in cells rely on a dynamic interplay of proteins and other components. It has been an enormously productive approach, but it is also widely acknowledged now that in some systems the behaviour of interest is an emergent property that cannot be discerned in the separate parts. The reductionist approach to science seeks to understand the behaviour of systems by studying their individual components.

Molecules by Philip Ball