The Future of Ligand Engineering in Colloidal Semiconductor Nanocrystals

[Image: see text] Next-generation colloidal semiconductor nanocrystals featuring enhanced optoelectronic properties and processability are expected to arise from complete mastering of the nanocrystals’ surface characteristics, attained by a rational engineering of the passivating ligands. This aspect is highly challenging, as it underlies a detailed understanding of the critical chemical processes that occur at the nanocrystal–ligand–solvent interface, a task that is prohibitive because of the limited number of nanocrystal syntheses that could be tried in the lab, where only a few dozen of the commercially available starting ligands can actually be explored. However, this challenging goal can be addressed nowadays by combining experiments with atomistic calculations and machine learning algorithms. In the last decades we indeed witnessed major advances in the development and application of computational software dedicated to the solution of the electronic structure problem as well as the expansion of tools to improve the sampling and analysis in classical molecular dynamics simulations. More recently, this progress has also embraced the integration of machine learning in computational chemistry and in the discovery of new drugs. We expect that soon this plethora of computational tools will have a formidable impact also in the field of colloidal semiconductor nanocrystals. In this Account, we present some of the most recent developments in the atomistic description of colloidal nanocrystals. In particular, we show how our group has been developing a set of programs interfaced with available computational chemistry software packages that allow the thermodynamic controlling factors in the nanocrystal surface chemistry to be captured atomistically by including explicit solvent molecules, ligands, and nanocrystal sizes that match the experiments. At the same time, we are also setting up an infrastructure to automate the efficient execution of thousands of calculations that will enable the collection of sufficient data to be processed by machine learning. To fully capture the power of these computational tools in the chemistry of colloidal nanocrystals, we decided to embed the thermodynamics behind the dissolution/precipitation of nanocrystal–ligand complexes in organic solvents and the crucial process of binding/detachment of ligands at the nanocrystal surface into a unique chemical framework. We show that formalizing this mechanism with a computational bird’s eye view helps in deducing the critical factors that govern the stabilization of colloidal dispersions of nanocrystals in an organic solvent as well as the definition of those key parameters that need to be calculated to manipulate surface ligands. This approach has the ultimate goal of engineering surface ligands in silico, anticipating and driving the experiments in the lab.