The development of new, green processes requires concepts that enable the most efficient use of both energy and materials while eliminating environmental footprint. The most critical process operations, such as reactions and separations/purifications, often require optimal selection of process solvents and thermodynamic conditions. Ionic liquids form a broad class of solvents which can often be applied to develop green processes. However, the diversity of options for charge centers, chemical group functionality, valency and molecular configuration presents the most significant obstacle to their selection. In order to accelerate the development of green processes, a new molecular design methodology, COSMOdesign©, has been developed. COSMOdesign is an inverse methodology based on genetic Monte Carlo to find optimal compounds or chemical functionalities on molecular structures to satisfy bounds on thermodynamic properties. Given a process with a thermodynamic measure of efficiency and a set of underdeterminant process constraints, COSMOdesign will lead to an optimal selection of an ionic liquid mixture. Compounds are described using a conductor-like screening model and selective mutations are made on the polarization charges of a selected component. The evolution of mutations required to optimize the process thermodynamic efficiency enables the engineer to both assess the opportunity for improving the process design as well as identify the improved process components. Here the COSMOdesign methodology is described and applied to green process separations. Genetic Monte Carlo is applied to problems in which fixed solubilities are required at a fixed thermodynamic state and to different problems in which component partitioning must be enhanced by changing thermodynamic state. The genetic Monte Carlo methodology is also characterized for its efficiency to improve thermodynamic properties of mixtures, provide predictive understanding of ionic liquid mixtures and screen new application concepts.