The sunlight that floods our planet each day is the ultimate unlimited energy resource: it generates all of the carbon-based fuels and drives our food chain. Access to this power supply, however, critically depends upon catalysis, which is crucial to the photosynthetic conversion of light into organic matter1.
Catalysts work by modifying the energy barriers of chemical processes and accelerating reaction speeds without being consumed themselves. Normally, such processes are classified as one of the two types: homogenous if the catalyst is in the same phase (gas, liquid or solid) as the reactants, or heterogeneous if the phases are different. Many natural catalysts are homogenous; for example, the conversion of ozone into oxygen in the atmosphere is catalysed by chlorine radicals, all of which are in a gaseous phase. Heterogeneous systems commonly involve a solid metal or metal oxide catalyst interacting with gaseous reagents and can, for example, make possible the production of fertilizers or the reduction of pollution.
Catalysis research has entered a new age, marking the end of an era dominated by trial and error. Instead, sophisticated analytical tools and powerful computational methods are improving our understanding of catalysts under real conditions. Because the boundaries between homogenous catalysts, heterogeneous catalysts and biocatalysts are rapidly disappearing, breakthroughs will require contributions from multiple disciplines — making this a true scientific frontier.
Despite much progress, molecular-level descriptions of real catalyst systems are still rare. In particular, reaction rates are not easy to predict or measure; the selectivity observed in catalytic transformations depends on minute differences in such reaction rates. Moreover, the role of solvent molecules during liquid-phase homogenous reactions and the intricate dynamic properties of the catalysts themselves are still poorly understood.
New computational chemistry techniques can be used to calculate the structure and properties of molecules using quantum-mechanical theories2. Sophisticated methods that simulate catalysts at multiple timescales and over different lengths can be coupled with advanced computational power to help address many of these outstanding problems and generate predictive insights into new catalytic substances.