The fast development of digitalization and computational science is opening new possibilities for a rapid design of new materials. Computational tools coupled with focused experiments can be successfully used for the design of new nanostructured materials in different sectors, including membrane engineering. Accordingly, in this Chapter, we present the application of a multiscale molecular simulation protocol for predicting gas transport properties in polymeric nanocomposite membranes constituted by titania (TiO2) nanoparticles dispersed in thermoplastic polyurethanes (TPUs). The Chapter starts with a general introduction on the future of computational tools for the design of new materials and introduces the paradigm underlying of multiscale molecular modelling. It then continues with the description of the multiscale (i.e., atomistic, mesoscale and finite element calculations) computational recipe developed ad hoc for the prediction of different gas permeation and diffusion in TPU/TiO2 nanocomposite membranes. Finally, the comparison of in silico and experimental results on these systems is reported and discussed. The quality of the agreement obtained between virtual and real data for such complex systems indeed confirms the validity of computational tools for the design and transport property prediction of nanocomposite membranes for gas treatment.
