Multiscale modelling of photoinduced processes in composite systems

In the past few decades, quantum mechanical (QM) modelling has moved from isolated molecules made of few atoms to large supramolecular aggregates embedded in complex environments. The integration of QM methods within classical descriptions in multiscale models made such advances possible. One of the first examples of this integration is represented by continuum solvation models that have been largely and successfully applied to predict properties and processes of solvated molecules since the 1980s. Almost in the same years, an alternative classical description based on molecular mechanics (MM) was coupled to QM methods in hybrid QM/MM approaches. Since their first formulations, these QM/classical models have seen great development in terms of accuracy, robustness and generalizability. This progress has enabled their application to systems of increasing complexity and processes never studied before within a QM framework, such as photoinduced processes in biomolecules, nanomaterials and, more generally, composite systems. These systems bring together components of different sizes — molecular, nano and mesoscopic — and multiscale approaches enable their simultaneous investigation. In this Review, we highlight potentials and limitations of multiscale approaches for the modelling of photoinduced processes in composite systems. We discuss the developments that are still needed to elevate the QM-based multiscale strategy to a gold standard for the prediction of light-activated events in composite systems and the analysis of the outputs of novel advanced spectroscopies