Porous solids such as zeolites1 and metal–organic frameworks2, 3 are useful in molecular separation and in catalysis, but their solid nature can impose limitations. For example, liquid solvents, rather than porous solids, are the most mature technology for post-combustion capture of carbon dioxide because liquid circulation systems are more easily retrofitted to existing plants. Solid porous adsorbents offer major benefits, such as lower energy penalties in adsorption–desorption cycles4, but they are difficult to implement in conventional flow processes. Materials that combine the properties of fluidity and permanent porosity could therefore offer technological advantages, but permanent porosity is not associated with conventional liquids5. Here we report free-flowing liquids whose bulk properties are determined by their permanent porosity. To achieve this, we designed cage molecules6, 7 that provide a well-defined pore space and that are highly soluble in solvents whose molecules are too large to enter the pores. The concentration of unoccupied cages can thus be around 500 times greater than in other molecular solutions that contain cavities8, 9, 10, resulting in a marked change in bulk properties, such as an eightfold increase in the solubility of methane gas. Our results provide the basis for development of a new class of functional porous materials for chemical processes, and we present a one-step, multigram scale-up route for highly soluble ‘scrambled’ porous cages prepared from a mixture of commercially available reagents. The unifying design principle for these materials is the avoidance of functional groups that can penetrate into the molecular cage cavities.
a, Synthesis of the crown-ether cage. b, The empty, highly soluble cage molecule, left, defines the pore space; the 15-crown-5 solvent, middle, provides fluidity but cannot enter the cage cavities. The concentrated solution (porous liquid) flows at room temperature, right. Key: C, grey; O, red; N, blue; H, white. Space-filling rendering highlights the core of the cage. Ball and stick rendering represents the crown-ether substituents on the cage and the 15-crown-5 solvent. All H atoms except those attached to aromatic rings of the cage compound have been omitted for clarity.
Figure 2: Molecular simulations for the porous liquid show unoccupied molecular-sized pores.
a, Representative configuration of the porous liquid at 350 K. To highlight the cage cores, crown-ether solvent molecules and crown-ether substituents on the cages have been omitted. Red and yellow surfaces indicate empty pores inside or outside the cages, respectively. b, Relative porosity, Vrel(R), of the porous liquid at 350 K and 400 K. Inset, expansion of the 400 K result. At 350 K, the porous liquid has around 1,900 times as many methane-sized cavities (probe radius ~0.24 nm) than does the pure solvent.
Figure 3: Dissolution of methane in the porous liquid.
a, Methane is around 8 times more soluble per mass of the porous liquid at 1 atm pressure than per mass of the pure 15-crown-5 solvent. Error bars represent overall uncertainty of experimental data calculated by error propagation of the s.d. of the measured quantities. b, Molecular simulation of the porous liquid with dissolved methane (grand canonical Monte Carlo simulation at 350 K, 1 atm). Methane is shown in red when inside a cage core (<2.5 Å from the cage centre) or yellow when outside. Solvent and cage substituents are omitted for clarity.