Abstract
Rubbery organic frameworks-ROFs have recently emerged as an intriguing class of dynamers by virtue of reversible connections between their building units. Their highly adaptative features at the origin of their spectacular self-healing properties made them also attractive candidates for the development of gas-selective membranes combining high selectivity and fast permeability. So far, little is known on the origin of this unique trait and this clearly hampers the exploitation of this class of dynamers in many areas where stimuli-responsive pore dynamics is of great importance. To address this lack of fundamental knowledge, herein we unravel the self-assembly process of ROFs via the development of an advanced computational methodology combining quantum and force field molecular simulations that enable the description of reversible connections of building units and the long-range organization of the cross-linked ROF network. We demonstrate that both accurate energy barriers associated with the covalent bond formation between the building units and presence of solvent are key parameters to ensure the in silico construction of reliable ROF structure models that are supported by a set of experimental data collected on synthesized ROFs including density, connectivity and porosity. Atomistic insights into the unusual guest-responsive pore dynamics of this intriguing class of dynamers are further gained with a special attention paid to the tunability of this pore flexibility by controlling the chemical composition of the building units. As a further stage, the dynamics of CO(2) in these compliance frameworks is scrutinized to shed light on the mechanism at the origin of their promising performance as CO(2)-selective membranes. We highlight that guest-triggered pore dynamics enables the creation of a diffusion pathway to ensure effective gas transport throughout the whole ROF. This knowledge of the pore structure and its guest-responsive dynamics at the microscopic level is unprecedented in the field of dynamers and it is expected to pave the way towards the optimization of this class of adaptive porous frameworks for many potential applications. Interestingly, this computational approach can be transferable to the exploration of any complex disordered systems showing a high degree of flexibility and guest induced structure/pore reorganization.