Computer-aided molecular and product design (CAMPD) can in principle be used to find simultaneously
the optimal conditions in separation processes and the structure of the optimal solvents.
In many cases, however, the solution of CAMPD problems is challenging. In this paper, we propose
a solution approach for the CAMPD of solvent-based separation systems in which implicit
constraints on phase behaviour in process models are used to test the feasibility of the process
and solvent domains. The tests not only eliminate infeasible molecules from the search space but
also infeasible combinations of solvent molecules and process conditions. The tests also provide
bounds for the optimization of the process model (primal problem) for each solvent, facilitating
numerical solution. This is demonstrated on a prototypical natural gas purification process.