Challenges and opportunities of molecular simulations for negative gas adsorption

Negative gas adsorption (NGA) is a particularly eye-catching phenomenon, involving the spontaneous desorption of gas upon pressure increase during adsorption in a flexible nanoporous material. The material undergoes a structural transition from an “open pore” phase to a contracted “closed pore” phase upon gas adsorption, leading to macroscopic gas desorption visible to the naked eye. It was initially evidenced experimentally in 2016 for the adsorption of methane and n-butane in the DUT-49 metal–organic framework (DUT = Dresden University of Technology), and later demonstrated to be a general phenomenon, occurring for different gases and in a variety of materials with the same topology. NGA materials belong to the category of metamaterials, displaying behavior that is not found (or rarely observed) in “natural” or simple materials. The negative adsorption transition takes place outside of thermodynamic equilibrium, and its characterization requires the use of many complementary experimental techniques (adsorption measurements, in situ X-ray diffraction, EXAFS, NMR, etc.), as well as molecular simulation techniques. In order to obtain a full and consistent picture of the NGA phenomenon, it is indeed necessary to combine computational modelling with a variety of methods, at different scales, in order to understand the microscopic behavior of the host framework and guest molecules to the macroscopic experimental results. At the smallest scale, density functional theory (DFT) calculations have been used to understand the energetics and structure of the NGA materials, as well as the micromechanical properties of their organic linkers: the buckling of these linkers explains the large metastability of the open-pore phase, and gives rise to the NGA transition. At a larger scale, classical grand canonical Monte Carlo (GCMC) simulations in the “rigid host” structures can predict the adsorption capacity of different phases, elucidating the driving force behind the structural transition. To explicitly couple the flexibility of the framework and the adsorption of guest molecules, molecular dynamics simulations (relying on a classical force field for the flexible MOF) can be coupled with free energy methods to investigate the thermodynamics of NGA, obtaining free energy profiles that determine the relative stability of different phases with varying amounts of adsorbed gas. Finally, mesoscopic-scale modeling methods are required in order to understand the phenomenon at a scale larger than one unit cell, and explain experimental findings about the influence of crystal size effects on the NGA transition. This Account summarizes the computational approaches that have been used so far to better understand negative gas adsorption, and highlight open questions and perspectives in this field of research.